{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 2.594611644744873, "learning_rate": 9e-07, "loss": 1.4305, "step": 10 }, { "grad_norm": 2.1219100952148438, "learning_rate": 1.9e-06, "loss": 1.3391, "step": 20 }, { "grad_norm": 2.5379836559295654, "learning_rate": 2.9e-06, "loss": 1.3961, "step": 30 }, { "grad_norm": 1.303380012512207, "learning_rate": 3.9e-06, "loss": 1.3844, "step": 40 }, { "grad_norm": 1.0098382234573364, "learning_rate": 4.9000000000000005e-06, "loss": 1.3922, "step": 50 }, { "grad_norm": 1.0325270891189575, "learning_rate": 5.9e-06, "loss": 1.3313, "step": 60 }, { "grad_norm": 1.4544047117233276, "learning_rate": 6.900000000000001e-06, "loss": 1.318, "step": 70 }, { "grad_norm": 2.706803560256958, "learning_rate": 7.9e-06, "loss": 1.2398, "step": 80 }, { "grad_norm": 1.4012260437011719, "learning_rate": 8.9e-06, "loss": 1.2469, "step": 90 }, { "grad_norm": 1.9529749155044556, "learning_rate": 9.900000000000002e-06, "loss": 1.2109, "step": 100 }, { "grad_norm": 1.0739604234695435, "learning_rate": 1.09e-05, "loss": 1.1656, "step": 110 }, { "grad_norm": 1.2750731706619263, "learning_rate": 1.19e-05, "loss": 1.1062, "step": 120 }, { "grad_norm": 1.5115655660629272, "learning_rate": 1.29e-05, "loss": 1.1062, "step": 130 }, { "grad_norm": 1.7575154304504395, "learning_rate": 1.3900000000000002e-05, "loss": 1.1031, "step": 140 }, { "grad_norm": 1.7713111639022827, "learning_rate": 1.49e-05, "loss": 1.0898, "step": 150 }, { "grad_norm": 1.5832326412200928, "learning_rate": 1.59e-05, "loss": 1.0809, "step": 160 }, { "grad_norm": 1.6050209999084473, "learning_rate": 1.69e-05, "loss": 1.0609, "step": 170 }, { "grad_norm": 1.4061970710754395, "learning_rate": 1.79e-05, "loss": 1.0977, "step": 180 }, { "grad_norm": 1.5004452466964722, "learning_rate": 1.8900000000000002e-05, "loss": 1.0566, "step": 190 }, { "grad_norm": 1.2839986085891724, "learning_rate": 1.9900000000000003e-05, "loss": 1.0641, "step": 200 }, { "grad_norm": 1.1661006212234497, "learning_rate": 2.09e-05, "loss": 1.0492, "step": 210 }, { "grad_norm": 1.0299561023712158, "learning_rate": 2.19e-05, "loss": 1.0434, "step": 220 }, { "grad_norm": 1.1774638891220093, "learning_rate": 2.29e-05, "loss": 1.0324, "step": 230 }, { "grad_norm": 1.413106918334961, "learning_rate": 2.39e-05, "loss": 1.0508, "step": 240 }, { "grad_norm": 1.543010950088501, "learning_rate": 2.4900000000000002e-05, "loss": 1.0469, "step": 250 }, { "grad_norm": 1.3146458864212036, "learning_rate": 2.5900000000000003e-05, "loss": 1.0289, "step": 260 }, { "grad_norm": 1.1733176708221436, "learning_rate": 2.6900000000000003e-05, "loss": 1.0523, "step": 270 }, { "grad_norm": 0.9241418242454529, "learning_rate": 2.7900000000000004e-05, "loss": 1.0715, "step": 280 }, { "grad_norm": 1.4385899305343628, "learning_rate": 2.8899999999999998e-05, "loss": 1.0387, "step": 290 }, { "grad_norm": 2.058316469192505, "learning_rate": 2.9900000000000002e-05, "loss": 1.0242, "step": 300 }, { "grad_norm": 1.5661393404006958, "learning_rate": 3.09e-05, "loss": 1.0336, "step": 310 }, { "grad_norm": 1.5798689126968384, "learning_rate": 3.19e-05, "loss": 0.9969, "step": 320 }, { "grad_norm": 1.3412009477615356, "learning_rate": 3.29e-05, "loss": 0.9945, "step": 330 }, { "grad_norm": 1.0843878984451294, "learning_rate": 3.3900000000000004e-05, "loss": 0.9836, "step": 340 }, { "grad_norm": 1.239405870437622, "learning_rate": 3.49e-05, "loss": 0.9637, "step": 350 }, { "grad_norm": 1.2260103225708008, "learning_rate": 3.59e-05, "loss": 0.9832, "step": 360 }, { "grad_norm": 1.4297655820846558, "learning_rate": 3.69e-05, "loss": 0.9402, "step": 370 }, { "grad_norm": 1.01796293258667, "learning_rate": 3.79e-05, "loss": 0.9547, "step": 380 }, { "grad_norm": 1.444282054901123, "learning_rate": 3.8900000000000004e-05, "loss": 0.9762, "step": 390 }, { "grad_norm": 1.7160881757736206, "learning_rate": 3.99e-05, "loss": 0.9215, "step": 400 }, { "grad_norm": 1.986338496208191, "learning_rate": 4.09e-05, "loss": 0.9062, "step": 410 }, { "grad_norm": 3.2725255489349365, "learning_rate": 4.19e-05, "loss": 0.8926, "step": 420 }, { "grad_norm": 2.4951891899108887, "learning_rate": 4.29e-05, "loss": 0.8738, "step": 430 }, { "grad_norm": 2.2020833492279053, "learning_rate": 4.39e-05, "loss": 0.8391, "step": 440 }, { "grad_norm": 2.768505573272705, "learning_rate": 4.49e-05, "loss": 0.8398, "step": 450 }, { "grad_norm": 1.5553258657455444, "learning_rate": 4.5900000000000004e-05, "loss": 0.7926, "step": 460 }, { "grad_norm": 1.8189395666122437, "learning_rate": 4.69e-05, "loss": 0.782, "step": 470 }, { "grad_norm": 1.746628761291504, "learning_rate": 4.79e-05, "loss": 0.741, "step": 480 }, { "grad_norm": 1.644413709640503, "learning_rate": 4.89e-05, "loss": 0.6816, "step": 490 }, { "grad_norm": 1.784613847732544, "learning_rate": 4.99e-05, "loss": 0.6598, "step": 500 }, { "grad_norm": 1.3432937860488892, "learning_rate": 5.0900000000000004e-05, "loss": 0.6203, "step": 510 }, { "grad_norm": 1.4880379438400269, "learning_rate": 5.19e-05, "loss": 0.5932, "step": 520 }, { "grad_norm": 1.331079363822937, "learning_rate": 5.2900000000000005e-05, "loss": 0.5545, "step": 530 }, { "grad_norm": 1.9008127450942993, "learning_rate": 5.390000000000001e-05, "loss": 0.4949, "step": 540 }, { "grad_norm": 1.9666775465011597, "learning_rate": 5.4900000000000006e-05, "loss": 0.5449, "step": 550 }, { "grad_norm": 1.4349759817123413, "learning_rate": 5.590000000000001e-05, "loss": 0.4967, "step": 560 }, { "grad_norm": 1.4970329999923706, "learning_rate": 5.69e-05, "loss": 0.4381, "step": 570 }, { "grad_norm": 1.4168415069580078, "learning_rate": 5.79e-05, "loss": 0.4453, "step": 580 }, { "grad_norm": 2.2595198154449463, "learning_rate": 5.89e-05, "loss": 0.4029, "step": 590 }, { "grad_norm": 1.6374934911727905, "learning_rate": 5.99e-05, "loss": 0.366, "step": 600 }, { "grad_norm": 2.463252305984497, "learning_rate": 6.09e-05, "loss": 0.2735, "step": 610 }, { "grad_norm": 1.5838596820831299, "learning_rate": 6.19e-05, "loss": 0.249, "step": 620 }, { "grad_norm": 1.3732593059539795, "learning_rate": 6.29e-05, "loss": 0.267, "step": 630 }, { "grad_norm": 1.9342843294143677, "learning_rate": 6.390000000000001e-05, "loss": 0.2396, "step": 640 }, { "grad_norm": 2.1176393032073975, "learning_rate": 6.49e-05, "loss": 0.224, "step": 650 }, { "grad_norm": 2.224191188812256, "learning_rate": 6.59e-05, "loss": 0.2367, "step": 660 }, { "grad_norm": 1.2918319702148438, "learning_rate": 6.690000000000001e-05, "loss": 0.1897, "step": 670 }, { "grad_norm": 2.011927843093872, "learning_rate": 6.790000000000001e-05, "loss": 0.2027, "step": 680 }, { "grad_norm": 1.5603440999984741, "learning_rate": 6.89e-05, "loss": 0.2168, "step": 690 }, { "grad_norm": 1.7028965950012207, "learning_rate": 6.99e-05, "loss": 0.253, "step": 700 }, { "grad_norm": 1.8830788135528564, "learning_rate": 7.09e-05, "loss": 0.2013, "step": 710 }, { "grad_norm": 1.8330340385437012, "learning_rate": 7.19e-05, "loss": 0.2168, "step": 720 }, { "grad_norm": 1.1798551082611084, "learning_rate": 7.29e-05, "loss": 0.1911, "step": 730 }, { "grad_norm": 1.2433335781097412, "learning_rate": 7.390000000000001e-05, "loss": 0.1423, "step": 740 }, { "grad_norm": 2.1076877117156982, "learning_rate": 7.49e-05, "loss": 0.1577, "step": 750 }, { "grad_norm": 1.2687638998031616, "learning_rate": 7.59e-05, "loss": 0.1618, "step": 760 }, { "grad_norm": 1.7713589668273926, "learning_rate": 7.69e-05, "loss": 0.1805, "step": 770 }, { "grad_norm": 1.5923417806625366, "learning_rate": 7.790000000000001e-05, "loss": 0.2124, "step": 780 }, { "grad_norm": 1.4625893831253052, "learning_rate": 7.890000000000001e-05, "loss": 0.1567, "step": 790 }, { "grad_norm": 1.1644519567489624, "learning_rate": 7.99e-05, "loss": 0.148, "step": 800 }, { "grad_norm": 1.1727776527404785, "learning_rate": 8.090000000000001e-05, "loss": 0.1728, "step": 810 }, { "grad_norm": 1.2896478176116943, "learning_rate": 8.19e-05, "loss": 0.1795, "step": 820 }, { "grad_norm": 1.3555573225021362, "learning_rate": 8.29e-05, "loss": 0.2244, "step": 830 }, { "grad_norm": 1.1761329174041748, "learning_rate": 8.39e-05, "loss": 0.1651, "step": 840 }, { "grad_norm": 1.3670698404312134, "learning_rate": 8.49e-05, "loss": 0.1655, "step": 850 }, { "grad_norm": 1.376820683479309, "learning_rate": 8.59e-05, "loss": 0.148, "step": 860 }, { "grad_norm": 1.5688972473144531, "learning_rate": 8.69e-05, "loss": 0.166, "step": 870 }, { "grad_norm": 1.4234598875045776, "learning_rate": 8.790000000000001e-05, "loss": 0.137, "step": 880 }, { "grad_norm": 1.157599687576294, "learning_rate": 8.89e-05, "loss": 0.1365, "step": 890 }, { "grad_norm": 1.2716339826583862, "learning_rate": 8.99e-05, "loss": 0.1622, "step": 900 }, { "grad_norm": 1.9193576574325562, "learning_rate": 9.090000000000001e-05, "loss": 0.1811, "step": 910 }, { "grad_norm": 1.3307154178619385, "learning_rate": 9.190000000000001e-05, "loss": 0.1396, "step": 920 }, { "grad_norm": 1.3856769800186157, "learning_rate": 9.290000000000001e-05, "loss": 0.1481, "step": 930 }, { "grad_norm": 2.0725173950195312, "learning_rate": 9.39e-05, "loss": 0.1733, "step": 940 }, { "grad_norm": 0.9583559632301331, "learning_rate": 9.49e-05, "loss": 0.1238, "step": 950 }, { "grad_norm": 0.9393100142478943, "learning_rate": 9.59e-05, "loss": 0.1456, "step": 960 }, { "grad_norm": 1.11801016330719, "learning_rate": 9.69e-05, "loss": 0.1713, "step": 970 }, { "grad_norm": 0.9369545578956604, "learning_rate": 9.790000000000001e-05, "loss": 0.11, "step": 980 }, { "grad_norm": 1.4491400718688965, "learning_rate": 9.89e-05, "loss": 0.12, "step": 990 }, { "grad_norm": 1.686273217201233, "learning_rate": 9.99e-05, "loss": 0.1171, "step": 1000 }, { "grad_norm": 1.2652745246887207, "learning_rate": 9.999994463727085e-05, "loss": 0.1416, "step": 1010 }, { "grad_norm": 1.440425992012024, "learning_rate": 9.999975326009292e-05, "loss": 0.1571, "step": 1020 }, { "grad_norm": 1.2276880741119385, "learning_rate": 9.999942518549879e-05, "loss": 0.1146, "step": 1030 }, { "grad_norm": 0.8655964732170105, "learning_rate": 9.999896041438544e-05, "loss": 0.1095, "step": 1040 }, { "grad_norm": 0.7947432398796082, "learning_rate": 9.999835894802353e-05, "loss": 0.132, "step": 1050 }, { "grad_norm": 1.2246838808059692, "learning_rate": 9.999762078805743e-05, "loss": 0.1412, "step": 1060 }, { "grad_norm": 2.208711862564087, "learning_rate": 9.999674593650526e-05, "loss": 0.1495, "step": 1070 }, { "grad_norm": 1.51913321018219, "learning_rate": 9.99957343957588e-05, "loss": 0.1797, "step": 1080 }, { "grad_norm": 1.0692880153656006, "learning_rate": 9.99945861685836e-05, "loss": 0.1181, "step": 1090 }, { "grad_norm": 1.2480826377868652, "learning_rate": 9.999330125811884e-05, "loss": 0.1035, "step": 1100 }, { "grad_norm": 1.1540807485580444, "learning_rate": 9.999187966787744e-05, "loss": 0.1864, "step": 1110 }, { "grad_norm": 0.6933703422546387, "learning_rate": 9.999032140174595e-05, "loss": 0.106, "step": 1120 }, { "grad_norm": 0.989782452583313, "learning_rate": 9.998862646398464e-05, "loss": 0.1381, "step": 1130 }, { "grad_norm": 1.253698468208313, "learning_rate": 9.998679485922739e-05, "loss": 0.1959, "step": 1140 }, { "grad_norm": 0.9849686622619629, "learning_rate": 9.998482659248174e-05, "loss": 0.1459, "step": 1150 }, { "grad_norm": 1.007449746131897, "learning_rate": 9.998272166912883e-05, "loss": 0.1241, "step": 1160 }, { "grad_norm": 1.1079837083816528, "learning_rate": 9.998048009492347e-05, "loss": 0.1169, "step": 1170 }, { "grad_norm": 0.7529625296592712, "learning_rate": 9.997810187599403e-05, "loss": 0.1048, "step": 1180 }, { "grad_norm": 1.2914284467697144, "learning_rate": 9.997558701884249e-05, "loss": 0.1091, "step": 1190 }, { "grad_norm": 1.6250420808792114, "learning_rate": 9.997293553034433e-05, "loss": 0.1389, "step": 1200 }, { "grad_norm": 1.5275187492370605, "learning_rate": 9.997014741774866e-05, "loss": 0.1597, "step": 1210 }, { "grad_norm": 0.7246806621551514, "learning_rate": 9.996722268867803e-05, "loss": 0.1304, "step": 1220 }, { "grad_norm": 1.039007306098938, "learning_rate": 9.996416135112858e-05, "loss": 0.1199, "step": 1230 }, { "grad_norm": 0.77347332239151, "learning_rate": 9.996096341346988e-05, "loss": 0.1347, "step": 1240 }, { "grad_norm": 0.5832141637802124, "learning_rate": 9.995762888444495e-05, "loss": 0.1234, "step": 1250 }, { "grad_norm": 1.0936914682388306, "learning_rate": 9.995415777317027e-05, "loss": 0.1133, "step": 1260 }, { "grad_norm": 1.025694489479065, "learning_rate": 9.995055008913574e-05, "loss": 0.1086, "step": 1270 }, { "grad_norm": 1.2594245672225952, "learning_rate": 9.994680584220463e-05, "loss": 0.1289, "step": 1280 }, { "grad_norm": 0.7362596988677979, "learning_rate": 9.994292504261355e-05, "loss": 0.1129, "step": 1290 }, { "grad_norm": 1.0678772926330566, "learning_rate": 9.993890770097247e-05, "loss": 0.1113, "step": 1300 }, { "grad_norm": 0.7894284725189209, "learning_rate": 9.993475382826467e-05, "loss": 0.1028, "step": 1310 }, { "grad_norm": 1.253738284111023, "learning_rate": 9.993046343584664e-05, "loss": 0.1127, "step": 1320 }, { "grad_norm": 0.8095929622650146, "learning_rate": 9.992603653544816e-05, "loss": 0.0762, "step": 1330 }, { "grad_norm": 1.1425012350082397, "learning_rate": 9.992147313917222e-05, "loss": 0.0829, "step": 1340 }, { "grad_norm": 0.6637998819351196, "learning_rate": 9.991677325949497e-05, "loss": 0.1276, "step": 1350 }, { "grad_norm": 1.0939197540283203, "learning_rate": 9.991193690926568e-05, "loss": 0.1136, "step": 1360 }, { "grad_norm": 1.016067624092102, "learning_rate": 9.990696410170678e-05, "loss": 0.086, "step": 1370 }, { "grad_norm": 1.071104884147644, "learning_rate": 9.990185485041371e-05, "loss": 0.0825, "step": 1380 }, { "grad_norm": 0.7317979335784912, "learning_rate": 9.989660916935498e-05, "loss": 0.1137, "step": 1390 }, { "grad_norm": 0.9569719433784485, "learning_rate": 9.989122707287208e-05, "loss": 0.0802, "step": 1400 }, { "grad_norm": 1.1662852764129639, "learning_rate": 9.988570857567945e-05, "loss": 0.1312, "step": 1410 }, { "grad_norm": 0.9923275709152222, "learning_rate": 9.988005369286446e-05, "loss": 0.1192, "step": 1420 }, { "grad_norm": 1.7941197156906128, "learning_rate": 9.987426243988734e-05, "loss": 0.1057, "step": 1430 }, { "grad_norm": 0.8702324628829956, "learning_rate": 9.986833483258114e-05, "loss": 0.1084, "step": 1440 }, { "grad_norm": 1.1422405242919922, "learning_rate": 9.986227088715173e-05, "loss": 0.1106, "step": 1450 }, { "grad_norm": 0.9349690079689026, "learning_rate": 9.98560706201777e-05, "loss": 0.1001, "step": 1460 }, { "grad_norm": 0.8552647233009338, "learning_rate": 9.984973404861036e-05, "loss": 0.1023, "step": 1470 }, { "grad_norm": 1.005336880683899, "learning_rate": 9.984326118977361e-05, "loss": 0.0955, "step": 1480 }, { "grad_norm": 0.9886020421981812, "learning_rate": 9.983665206136406e-05, "loss": 0.1043, "step": 1490 }, { "grad_norm": 1.0652482509613037, "learning_rate": 9.982990668145075e-05, "loss": 0.1274, "step": 1500 }, { "grad_norm": 1.0435166358947754, "learning_rate": 9.982302506847534e-05, "loss": 0.0869, "step": 1510 }, { "grad_norm": 0.98117595911026, "learning_rate": 9.981600724125189e-05, "loss": 0.1013, "step": 1520 }, { "grad_norm": 0.8921933770179749, "learning_rate": 9.980885321896685e-05, "loss": 0.1012, "step": 1530 }, { "grad_norm": 0.5732268691062927, "learning_rate": 9.980156302117905e-05, "loss": 0.1057, "step": 1540 }, { "grad_norm": 0.8019578456878662, "learning_rate": 9.979413666781963e-05, "loss": 0.0857, "step": 1550 }, { "grad_norm": 0.8753455877304077, "learning_rate": 9.978657417919193e-05, "loss": 0.1049, "step": 1560 }, { "grad_norm": 0.7851360440254211, "learning_rate": 9.977887557597153e-05, "loss": 0.0781, "step": 1570 }, { "grad_norm": 0.9559298157691956, "learning_rate": 9.97710408792061e-05, "loss": 0.0864, "step": 1580 }, { "grad_norm": 1.5202651023864746, "learning_rate": 9.976307011031542e-05, "loss": 0.1237, "step": 1590 }, { "grad_norm": 0.8220219612121582, "learning_rate": 9.975496329109126e-05, "loss": 0.0902, "step": 1600 }, { "grad_norm": 0.8519560098648071, "learning_rate": 9.974672044369732e-05, "loss": 0.0703, "step": 1610 }, { "grad_norm": 0.8918028473854065, "learning_rate": 9.97383415906693e-05, "loss": 0.0921, "step": 1620 }, { "grad_norm": 0.5760334730148315, "learning_rate": 9.97298267549146e-05, "loss": 0.1118, "step": 1630 }, { "grad_norm": 0.7157757878303528, "learning_rate": 9.972117595971249e-05, "loss": 0.0971, "step": 1640 }, { "grad_norm": 0.8736289143562317, "learning_rate": 9.971238922871391e-05, "loss": 0.0751, "step": 1650 }, { "grad_norm": 0.8117972612380981, "learning_rate": 9.970346658594142e-05, "loss": 0.085, "step": 1660 }, { "grad_norm": 0.5310317277908325, "learning_rate": 9.969440805578923e-05, "loss": 0.0752, "step": 1670 }, { "grad_norm": 1.1288162469863892, "learning_rate": 9.968521366302298e-05, "loss": 0.0798, "step": 1680 }, { "grad_norm": 0.6637669205665588, "learning_rate": 9.967588343277981e-05, "loss": 0.0878, "step": 1690 }, { "grad_norm": 0.7950348854064941, "learning_rate": 9.966641739056818e-05, "loss": 0.0969, "step": 1700 }, { "grad_norm": 0.8237141966819763, "learning_rate": 9.965681556226793e-05, "loss": 0.1256, "step": 1710 }, { "grad_norm": 1.187766194343567, "learning_rate": 9.964707797413006e-05, "loss": 0.1055, "step": 1720 }, { "grad_norm": 0.9922911524772644, "learning_rate": 9.963720465277679e-05, "loss": 0.0845, "step": 1730 }, { "grad_norm": 0.8277208209037781, "learning_rate": 9.96271956252014e-05, "loss": 0.1144, "step": 1740 }, { "grad_norm": 0.9819006323814392, "learning_rate": 9.961705091876816e-05, "loss": 0.1088, "step": 1750 }, { "grad_norm": 1.074457049369812, "learning_rate": 9.960677056121235e-05, "loss": 0.1001, "step": 1760 }, { "grad_norm": 0.5798190236091614, "learning_rate": 9.959635458064005e-05, "loss": 0.0761, "step": 1770 }, { "grad_norm": 1.0080076456069946, "learning_rate": 9.958580300552815e-05, "loss": 0.1147, "step": 1780 }, { "grad_norm": 0.8988767266273499, "learning_rate": 9.957511586472426e-05, "loss": 0.0746, "step": 1790 }, { "grad_norm": 0.7422889471054077, "learning_rate": 9.956429318744662e-05, "loss": 0.0951, "step": 1800 }, { "grad_norm": 1.015955924987793, "learning_rate": 9.955333500328404e-05, "loss": 0.0779, "step": 1810 }, { "grad_norm": 0.8055348992347717, "learning_rate": 9.95422413421957e-05, "loss": 0.0851, "step": 1820 }, { "grad_norm": 1.5066837072372437, "learning_rate": 9.953101223451133e-05, "loss": 0.0993, "step": 1830 }, { "grad_norm": 1.229966163635254, "learning_rate": 9.951964771093085e-05, "loss": 0.0932, "step": 1840 }, { "grad_norm": 0.6262849569320679, "learning_rate": 9.950814780252442e-05, "loss": 0.1095, "step": 1850 }, { "grad_norm": 0.8715920448303223, "learning_rate": 9.949651254073236e-05, "loss": 0.1121, "step": 1860 }, { "grad_norm": 1.1337716579437256, "learning_rate": 9.948474195736504e-05, "loss": 0.0869, "step": 1870 }, { "grad_norm": 0.7262267470359802, "learning_rate": 9.947283608460277e-05, "loss": 0.0656, "step": 1880 }, { "grad_norm": 0.849579930305481, "learning_rate": 9.946079495499577e-05, "loss": 0.0794, "step": 1890 }, { "grad_norm": 0.715893030166626, "learning_rate": 9.944861860146401e-05, "loss": 0.0909, "step": 1900 }, { "grad_norm": 0.8535832166671753, "learning_rate": 9.943630705729719e-05, "loss": 0.0898, "step": 1910 }, { "grad_norm": 1.1465808153152466, "learning_rate": 9.942386035615459e-05, "loss": 0.0714, "step": 1920 }, { "grad_norm": 0.737372100353241, "learning_rate": 9.941127853206503e-05, "loss": 0.0711, "step": 1930 }, { "grad_norm": 1.133686900138855, "learning_rate": 9.939856161942673e-05, "loss": 0.1118, "step": 1940 }, { "grad_norm": 0.5814546942710876, "learning_rate": 9.938570965300724e-05, "loss": 0.0699, "step": 1950 }, { "grad_norm": 0.8902150392532349, "learning_rate": 9.937272266794335e-05, "loss": 0.1077, "step": 1960 }, { "grad_norm": 0.7570468187332153, "learning_rate": 9.935960069974096e-05, "loss": 0.0682, "step": 1970 }, { "grad_norm": 0.9781492948532104, "learning_rate": 9.934634378427506e-05, "loss": 0.0886, "step": 1980 }, { "grad_norm": 1.0472445487976074, "learning_rate": 9.933295195778954e-05, "loss": 0.0781, "step": 1990 }, { "grad_norm": 1.068117380142212, "learning_rate": 9.931942525689715e-05, "loss": 0.0877, "step": 2000 }, { "grad_norm": 1.076836347579956, "learning_rate": 9.930576371857936e-05, "loss": 0.0991, "step": 2010 }, { "grad_norm": 0.813343346118927, "learning_rate": 9.929196738018629e-05, "loss": 0.0744, "step": 2020 }, { "grad_norm": 1.1044321060180664, "learning_rate": 9.927803627943662e-05, "loss": 0.0594, "step": 2030 }, { "grad_norm": 0.8297349214553833, "learning_rate": 9.926397045441744e-05, "loss": 0.0835, "step": 2040 }, { "grad_norm": 0.7494553923606873, "learning_rate": 9.924976994358417e-05, "loss": 0.0985, "step": 2050 }, { "grad_norm": 0.730701208114624, "learning_rate": 9.923543478576048e-05, "loss": 0.0626, "step": 2060 }, { "grad_norm": 0.8480241298675537, "learning_rate": 9.922096502013813e-05, "loss": 0.0724, "step": 2070 }, { "grad_norm": 0.8357149958610535, "learning_rate": 9.92063606862769e-05, "loss": 0.0834, "step": 2080 }, { "grad_norm": 0.7468804717063904, "learning_rate": 9.919162182410453e-05, "loss": 0.0649, "step": 2090 }, { "grad_norm": 1.0187464952468872, "learning_rate": 9.917674847391645e-05, "loss": 0.0755, "step": 2100 }, { "grad_norm": 0.5649673342704773, "learning_rate": 9.916174067637584e-05, "loss": 0.0926, "step": 2110 }, { "grad_norm": 0.9315624833106995, "learning_rate": 9.914659847251348e-05, "loss": 0.0802, "step": 2120 }, { "grad_norm": 0.4495192766189575, "learning_rate": 9.913132190372753e-05, "loss": 0.0685, "step": 2130 }, { "grad_norm": 1.1884993314743042, "learning_rate": 9.911591101178359e-05, "loss": 0.0578, "step": 2140 }, { "grad_norm": 0.758231520652771, "learning_rate": 9.910036583881443e-05, "loss": 0.0997, "step": 2150 }, { "grad_norm": 0.6936259269714355, "learning_rate": 9.908468642731995e-05, "loss": 0.07, "step": 2160 }, { "grad_norm": 0.5645713210105896, "learning_rate": 9.906887282016707e-05, "loss": 0.0677, "step": 2170 }, { "grad_norm": 0.5568066835403442, "learning_rate": 9.90529250605896e-05, "loss": 0.076, "step": 2180 }, { "grad_norm": 0.7924133539199829, "learning_rate": 9.903684319218809e-05, "loss": 0.0769, "step": 2190 }, { "grad_norm": 0.6227429509162903, "learning_rate": 9.902062725892976e-05, "loss": 0.0805, "step": 2200 }, { "grad_norm": 0.7805032730102539, "learning_rate": 9.900427730514834e-05, "loss": 0.0752, "step": 2210 }, { "grad_norm": 0.5026907920837402, "learning_rate": 9.8987793375544e-05, "loss": 0.072, "step": 2220 }, { "grad_norm": 0.8788405656814575, "learning_rate": 9.897117551518318e-05, "loss": 0.0641, "step": 2230 }, { "grad_norm": 0.6426061391830444, "learning_rate": 9.895442376949844e-05, "loss": 0.0798, "step": 2240 }, { "grad_norm": 0.7785035371780396, "learning_rate": 9.893753818428845e-05, "loss": 0.0941, "step": 2250 }, { "grad_norm": 0.8681800365447998, "learning_rate": 9.892051880571773e-05, "loss": 0.0732, "step": 2260 }, { "grad_norm": 0.8744980096817017, "learning_rate": 9.890336568031663e-05, "loss": 0.0968, "step": 2270 }, { "grad_norm": 0.7377612590789795, "learning_rate": 9.888607885498113e-05, "loss": 0.0736, "step": 2280 }, { "grad_norm": 0.758029580116272, "learning_rate": 9.886865837697275e-05, "loss": 0.0988, "step": 2290 }, { "grad_norm": 0.5659455060958862, "learning_rate": 9.88511042939184e-05, "loss": 0.0762, "step": 2300 }, { "grad_norm": 0.9644096493721008, "learning_rate": 9.883341665381028e-05, "loss": 0.0786, "step": 2310 }, { "grad_norm": 0.8018860220909119, "learning_rate": 9.881559550500575e-05, "loss": 0.0617, "step": 2320 }, { "grad_norm": 0.577052116394043, "learning_rate": 9.879764089622712e-05, "loss": 0.0802, "step": 2330 }, { "grad_norm": 1.0776798725128174, "learning_rate": 9.87795528765616e-05, "loss": 0.0751, "step": 2340 }, { "grad_norm": 0.46561717987060547, "learning_rate": 9.876133149546118e-05, "loss": 0.0663, "step": 2350 }, { "grad_norm": 0.6238251328468323, "learning_rate": 9.874297680274238e-05, "loss": 0.0897, "step": 2360 }, { "grad_norm": 0.6339982748031616, "learning_rate": 9.872448884858624e-05, "loss": 0.0792, "step": 2370 }, { "grad_norm": 0.5500380992889404, "learning_rate": 9.870586768353815e-05, "loss": 0.0703, "step": 2380 }, { "grad_norm": 0.5550593137741089, "learning_rate": 9.868711335850764e-05, "loss": 0.0752, "step": 2390 }, { "grad_norm": 0.3631652593612671, "learning_rate": 9.866822592476833e-05, "loss": 0.0749, "step": 2400 }, { "grad_norm": 0.8651986122131348, "learning_rate": 9.86492054339577e-05, "loss": 0.0771, "step": 2410 }, { "grad_norm": 0.47137418389320374, "learning_rate": 9.863005193807711e-05, "loss": 0.0924, "step": 2420 }, { "grad_norm": 0.9145431518554688, "learning_rate": 9.861076548949143e-05, "loss": 0.0734, "step": 2430 }, { "grad_norm": 0.9869363903999329, "learning_rate": 9.859134614092912e-05, "loss": 0.0824, "step": 2440 }, { "grad_norm": 1.0175631046295166, "learning_rate": 9.857179394548191e-05, "loss": 0.0764, "step": 2450 }, { "grad_norm": 0.7275974750518799, "learning_rate": 9.855210895660477e-05, "loss": 0.0667, "step": 2460 }, { "grad_norm": 0.4651252329349518, "learning_rate": 9.853229122811568e-05, "loss": 0.063, "step": 2470 }, { "grad_norm": 0.49147942662239075, "learning_rate": 9.851234081419559e-05, "loss": 0.0705, "step": 2480 }, { "grad_norm": 1.226062297821045, "learning_rate": 9.849225776938814e-05, "loss": 0.0664, "step": 2490 }, { "grad_norm": 0.6819326281547546, "learning_rate": 9.847204214859964e-05, "loss": 0.0767, "step": 2500 }, { "grad_norm": 0.831283450126648, "learning_rate": 9.845169400709879e-05, "loss": 0.0731, "step": 2510 }, { "grad_norm": 0.789198100566864, "learning_rate": 9.843121340051664e-05, "loss": 0.0679, "step": 2520 }, { "grad_norm": 0.5620065331459045, "learning_rate": 9.841060038484641e-05, "loss": 0.0509, "step": 2530 }, { "grad_norm": 0.7675412893295288, "learning_rate": 9.838985501644328e-05, "loss": 0.0645, "step": 2540 }, { "grad_norm": 0.7205899953842163, "learning_rate": 9.83689773520243e-05, "loss": 0.0715, "step": 2550 }, { "grad_norm": 0.9214705228805542, "learning_rate": 9.834796744866819e-05, "loss": 0.0758, "step": 2560 }, { "grad_norm": 0.704910397529602, "learning_rate": 9.832682536381525e-05, "loss": 0.0803, "step": 2570 }, { "grad_norm": 0.5986775159835815, "learning_rate": 9.830555115526711e-05, "loss": 0.0645, "step": 2580 }, { "grad_norm": 0.652046799659729, "learning_rate": 9.828414488118667e-05, "loss": 0.0576, "step": 2590 }, { "grad_norm": 0.6977267265319824, "learning_rate": 9.826260660009785e-05, "loss": 0.07, "step": 2600 }, { "grad_norm": 1.0419113636016846, "learning_rate": 9.824093637088547e-05, "loss": 0.0808, "step": 2610 }, { "grad_norm": 0.6367712616920471, "learning_rate": 9.821913425279514e-05, "loss": 0.0552, "step": 2620 }, { "grad_norm": 0.7197160124778748, "learning_rate": 9.8197200305433e-05, "loss": 0.0697, "step": 2630 }, { "grad_norm": 0.685926079750061, "learning_rate": 9.817513458876564e-05, "loss": 0.0771, "step": 2640 }, { "grad_norm": 0.7553039789199829, "learning_rate": 9.815293716311987e-05, "loss": 0.0832, "step": 2650 }, { "grad_norm": 0.71351158618927, "learning_rate": 9.813060808918262e-05, "loss": 0.0764, "step": 2660 }, { "grad_norm": 0.5656139254570007, "learning_rate": 9.810814742800069e-05, "loss": 0.0858, "step": 2670 }, { "grad_norm": 0.6088170409202576, "learning_rate": 9.808555524098074e-05, "loss": 0.0885, "step": 2680 }, { "grad_norm": 0.6316095590591431, "learning_rate": 9.806283158988887e-05, "loss": 0.0682, "step": 2690 }, { "grad_norm": 0.43410202860832214, "learning_rate": 9.803997653685072e-05, "loss": 0.068, "step": 2700 }, { "grad_norm": 0.8787769079208374, "learning_rate": 9.801699014435112e-05, "loss": 0.069, "step": 2710 }, { "grad_norm": 0.7102867364883423, "learning_rate": 9.799387247523398e-05, "loss": 0.0784, "step": 2720 }, { "grad_norm": 0.713158369064331, "learning_rate": 9.797062359270215e-05, "loss": 0.071, "step": 2730 }, { "grad_norm": 0.5686906576156616, "learning_rate": 9.794724356031715e-05, "loss": 0.0636, "step": 2740 }, { "grad_norm": 0.7083339691162109, "learning_rate": 9.792373244199913e-05, "loss": 0.0774, "step": 2750 }, { "grad_norm": 0.7884225845336914, "learning_rate": 9.790009030202658e-05, "loss": 0.0607, "step": 2760 }, { "grad_norm": 0.4892207980155945, "learning_rate": 9.78763172050362e-05, "loss": 0.0552, "step": 2770 }, { "grad_norm": 1.1159330606460571, "learning_rate": 9.785241321602274e-05, "loss": 0.0896, "step": 2780 }, { "grad_norm": 0.3328188955783844, "learning_rate": 9.782837840033879e-05, "loss": 0.0794, "step": 2790 }, { "grad_norm": 0.614898681640625, "learning_rate": 9.780421282369461e-05, "loss": 0.0512, "step": 2800 }, { "grad_norm": 0.9833534359931946, "learning_rate": 9.777991655215797e-05, "loss": 0.103, "step": 2810 }, { "grad_norm": 0.5967358350753784, "learning_rate": 9.775548965215394e-05, "loss": 0.0667, "step": 2820 }, { "grad_norm": 0.5928962230682373, "learning_rate": 9.773093219046474e-05, "loss": 0.0613, "step": 2830 }, { "grad_norm": 0.8105804920196533, "learning_rate": 9.770624423422954e-05, "loss": 0.0761, "step": 2840 }, { "grad_norm": 0.8398619890213013, "learning_rate": 9.768142585094426e-05, "loss": 0.0885, "step": 2850 }, { "grad_norm": 1.158229947090149, "learning_rate": 9.765647710846142e-05, "loss": 0.0811, "step": 2860 }, { "grad_norm": 0.7034460306167603, "learning_rate": 9.763139807498991e-05, "loss": 0.0742, "step": 2870 }, { "grad_norm": 0.9154704809188843, "learning_rate": 9.760618881909487e-05, "loss": 0.0679, "step": 2880 }, { "grad_norm": 0.5531051158905029, "learning_rate": 9.758084940969744e-05, "loss": 0.0715, "step": 2890 }, { "grad_norm": 0.5503506660461426, "learning_rate": 9.755537991607459e-05, "loss": 0.0735, "step": 2900 }, { "grad_norm": 0.5193167328834534, "learning_rate": 9.752978040785895e-05, "loss": 0.0682, "step": 2910 }, { "grad_norm": 1.0523780584335327, "learning_rate": 9.750405095503859e-05, "loss": 0.0685, "step": 2920 }, { "grad_norm": 0.7663245797157288, "learning_rate": 9.747819162795686e-05, "loss": 0.0648, "step": 2930 }, { "grad_norm": 0.8352916836738586, "learning_rate": 9.745220249731217e-05, "loss": 0.0581, "step": 2940 }, { "grad_norm": 0.6768618822097778, "learning_rate": 9.742608363415781e-05, "loss": 0.0551, "step": 2950 }, { "grad_norm": 0.6908816695213318, "learning_rate": 9.739983510990176e-05, "loss": 0.072, "step": 2960 }, { "grad_norm": 0.45372119545936584, "learning_rate": 9.737345699630647e-05, "loss": 0.0635, "step": 2970 }, { "grad_norm": 0.6521028280258179, "learning_rate": 9.734694936548869e-05, "loss": 0.0642, "step": 2980 }, { "grad_norm": 0.414142370223999, "learning_rate": 9.732031228991932e-05, "loss": 0.0763, "step": 2990 }, { "grad_norm": 0.446503221988678, "learning_rate": 9.729354584242302e-05, "loss": 0.058, "step": 3000 }, { "grad_norm": 0.7834926247596741, "learning_rate": 9.726665009617832e-05, "loss": 0.0527, "step": 3010 }, { "grad_norm": 0.4211423397064209, "learning_rate": 9.723962512471714e-05, "loss": 0.0564, "step": 3020 }, { "grad_norm": 0.7177088856697083, "learning_rate": 9.72124710019247e-05, "loss": 0.0771, "step": 3030 }, { "grad_norm": 0.5042608380317688, "learning_rate": 9.718518780203934e-05, "loss": 0.0668, "step": 3040 }, { "grad_norm": 0.48655936121940613, "learning_rate": 9.715777559965228e-05, "loss": 0.0811, "step": 3050 }, { "grad_norm": 0.5464625954627991, "learning_rate": 9.713023446970746e-05, "loss": 0.0746, "step": 3060 }, { "grad_norm": 0.9344635605812073, "learning_rate": 9.710256448750126e-05, "loss": 0.0526, "step": 3070 }, { "grad_norm": 0.5958772897720337, "learning_rate": 9.707476572868235e-05, "loss": 0.0699, "step": 3080 }, { "grad_norm": 0.8779199719429016, "learning_rate": 9.704683826925149e-05, "loss": 0.0944, "step": 3090 }, { "grad_norm": 0.5800724625587463, "learning_rate": 9.701878218556129e-05, "loss": 0.0774, "step": 3100 }, { "grad_norm": 0.5204524397850037, "learning_rate": 9.699059755431598e-05, "loss": 0.0452, "step": 3110 }, { "grad_norm": 0.5162118673324585, "learning_rate": 9.696228445257132e-05, "loss": 0.0509, "step": 3120 }, { "grad_norm": 0.5358502268791199, "learning_rate": 9.693384295773419e-05, "loss": 0.049, "step": 3130 }, { "grad_norm": 0.5742571353912354, "learning_rate": 9.690527314756259e-05, "loss": 0.0547, "step": 3140 }, { "grad_norm": 0.7322394251823425, "learning_rate": 9.687657510016527e-05, "loss": 0.0579, "step": 3150 }, { "grad_norm": 0.6694306135177612, "learning_rate": 9.684774889400161e-05, "loss": 0.057, "step": 3160 }, { "grad_norm": 0.7136942148208618, "learning_rate": 9.681879460788135e-05, "loss": 0.0759, "step": 3170 }, { "grad_norm": 0.6505119800567627, "learning_rate": 9.67897123209644e-05, "loss": 0.0698, "step": 3180 }, { "grad_norm": 0.762866199016571, "learning_rate": 9.676050211276062e-05, "loss": 0.0703, "step": 3190 }, { "grad_norm": 0.6046313047409058, "learning_rate": 9.673116406312962e-05, "loss": 0.0582, "step": 3200 }, { "grad_norm": 0.5147132277488708, "learning_rate": 9.67016982522805e-05, "loss": 0.0532, "step": 3210 }, { "grad_norm": 0.7426005601882935, "learning_rate": 9.667210476077164e-05, "loss": 0.0808, "step": 3220 }, { "grad_norm": 0.6041998863220215, "learning_rate": 9.664238366951055e-05, "loss": 0.0744, "step": 3230 }, { "grad_norm": 0.6544203162193298, "learning_rate": 9.661253505975355e-05, "loss": 0.0593, "step": 3240 }, { "grad_norm": 0.5822626352310181, "learning_rate": 9.658255901310557e-05, "loss": 0.0544, "step": 3250 }, { "grad_norm": 0.5456928014755249, "learning_rate": 9.655245561152e-05, "loss": 0.0589, "step": 3260 }, { "grad_norm": 0.5643149614334106, "learning_rate": 9.65222249372984e-05, "loss": 0.0534, "step": 3270 }, { "grad_norm": 0.6426769495010376, "learning_rate": 9.649186707309026e-05, "loss": 0.0598, "step": 3280 }, { "grad_norm": 0.8299885392189026, "learning_rate": 9.646138210189283e-05, "loss": 0.0519, "step": 3290 }, { "grad_norm": 0.4046460688114166, "learning_rate": 9.643077010705087e-05, "loss": 0.0476, "step": 3300 }, { "grad_norm": 0.4814707040786743, "learning_rate": 9.640003117225637e-05, "loss": 0.0578, "step": 3310 }, { "grad_norm": 0.43306764960289, "learning_rate": 9.636916538154846e-05, "loss": 0.0735, "step": 3320 }, { "grad_norm": 0.6036322116851807, "learning_rate": 9.633817281931296e-05, "loss": 0.0633, "step": 3330 }, { "grad_norm": 0.7911640405654907, "learning_rate": 9.630705357028242e-05, "loss": 0.0681, "step": 3340 }, { "grad_norm": 0.7991747856140137, "learning_rate": 9.627580771953563e-05, "loss": 0.0683, "step": 3350 }, { "grad_norm": 0.6578499674797058, "learning_rate": 9.624443535249759e-05, "loss": 0.047, "step": 3360 }, { "grad_norm": 0.45319682359695435, "learning_rate": 9.621293655493913e-05, "loss": 0.0714, "step": 3370 }, { "grad_norm": 0.65953528881073, "learning_rate": 9.618131141297675e-05, "loss": 0.1099, "step": 3380 }, { "grad_norm": 0.5572097301483154, "learning_rate": 9.614956001307242e-05, "loss": 0.069, "step": 3390 }, { "grad_norm": 0.5348389148712158, "learning_rate": 9.611768244203321e-05, "loss": 0.0536, "step": 3400 }, { "grad_norm": 0.6513969898223877, "learning_rate": 9.60856787870112e-05, "loss": 0.0541, "step": 3410 }, { "grad_norm": 0.7358964085578918, "learning_rate": 9.605354913550318e-05, "loss": 0.0503, "step": 3420 }, { "grad_norm": 0.5238944292068481, "learning_rate": 9.602129357535037e-05, "loss": 0.0517, "step": 3430 }, { "grad_norm": 1.3322851657867432, "learning_rate": 9.598891219473825e-05, "loss": 0.0817, "step": 3440 }, { "grad_norm": 0.5026893615722656, "learning_rate": 9.595640508219625e-05, "loss": 0.0521, "step": 3450 }, { "grad_norm": 0.6546522378921509, "learning_rate": 9.592377232659761e-05, "loss": 0.054, "step": 3460 }, { "grad_norm": 0.5725192427635193, "learning_rate": 9.589101401715904e-05, "loss": 0.0737, "step": 3470 }, { "grad_norm": 0.5588870048522949, "learning_rate": 9.585813024344045e-05, "loss": 0.062, "step": 3480 }, { "grad_norm": 0.7879475355148315, "learning_rate": 9.58251210953449e-05, "loss": 0.0639, "step": 3490 }, { "grad_norm": 0.4204157888889313, "learning_rate": 9.579198666311809e-05, "loss": 0.0337, "step": 3500 }, { "grad_norm": 0.8428052067756653, "learning_rate": 9.575872703734832e-05, "loss": 0.0621, "step": 3510 }, { "grad_norm": 0.44087162613868713, "learning_rate": 9.572534230896611e-05, "loss": 0.0526, "step": 3520 }, { "grad_norm": 0.6228436231613159, "learning_rate": 9.569183256924403e-05, "loss": 0.0562, "step": 3530 }, { "grad_norm": 0.3012903034687042, "learning_rate": 9.565819790979646e-05, "loss": 0.0571, "step": 3540 }, { "grad_norm": 0.7498623132705688, "learning_rate": 9.562443842257925e-05, "loss": 0.0642, "step": 3550 }, { "grad_norm": 0.4117112457752228, "learning_rate": 9.559055419988956e-05, "loss": 0.0428, "step": 3560 }, { "grad_norm": 0.6686583757400513, "learning_rate": 9.555654533436557e-05, "loss": 0.0563, "step": 3570 }, { "grad_norm": 0.6271094679832458, "learning_rate": 9.552241191898621e-05, "loss": 0.0738, "step": 3580 }, { "grad_norm": 0.6279948353767395, "learning_rate": 9.548815404707092e-05, "loss": 0.061, "step": 3590 }, { "grad_norm": 0.5885175466537476, "learning_rate": 9.545377181227942e-05, "loss": 0.0594, "step": 3600 }, { "grad_norm": 0.605475127696991, "learning_rate": 9.541926530861145e-05, "loss": 0.0503, "step": 3610 }, { "grad_norm": 0.535317063331604, "learning_rate": 9.538463463040645e-05, "loss": 0.0475, "step": 3620 }, { "grad_norm": 0.4351852834224701, "learning_rate": 9.534987987234337e-05, "loss": 0.0368, "step": 3630 }, { "grad_norm": 0.43462803959846497, "learning_rate": 9.53150011294404e-05, "loss": 0.0466, "step": 3640 }, { "grad_norm": 0.5325431823730469, "learning_rate": 9.527999849705471e-05, "loss": 0.0509, "step": 3650 }, { "grad_norm": 0.4574865400791168, "learning_rate": 9.524487207088213e-05, "loss": 0.055, "step": 3660 }, { "grad_norm": 0.5633252859115601, "learning_rate": 9.520962194695698e-05, "loss": 0.0675, "step": 3670 }, { "grad_norm": 0.7528544664382935, "learning_rate": 9.517424822165175e-05, "loss": 0.0679, "step": 3680 }, { "grad_norm": 0.67596834897995, "learning_rate": 9.513875099167685e-05, "loss": 0.0477, "step": 3690 }, { "grad_norm": 0.4211680591106415, "learning_rate": 9.510313035408035e-05, "loss": 0.0468, "step": 3700 }, { "grad_norm": 0.7330886721611023, "learning_rate": 9.506738640624775e-05, "loss": 0.0517, "step": 3710 }, { "grad_norm": 0.43227821588516235, "learning_rate": 9.50315192459016e-05, "loss": 0.0588, "step": 3720 }, { "grad_norm": 0.3474389910697937, "learning_rate": 9.499552897110136e-05, "loss": 0.0566, "step": 3730 }, { "grad_norm": 0.6833894848823547, "learning_rate": 9.495941568024304e-05, "loss": 0.054, "step": 3740 }, { "grad_norm": 0.3972979485988617, "learning_rate": 9.492317947205904e-05, "loss": 0.0592, "step": 3750 }, { "grad_norm": 0.513540506362915, "learning_rate": 9.488682044561775e-05, "loss": 0.0419, "step": 3760 }, { "grad_norm": 0.6212592124938965, "learning_rate": 9.485033870032335e-05, "loss": 0.0738, "step": 3770 }, { "grad_norm": 0.6534880995750427, "learning_rate": 9.481373433591556e-05, "loss": 0.0573, "step": 3780 }, { "grad_norm": 0.5454269051551819, "learning_rate": 9.47770074524693e-05, "loss": 0.0509, "step": 3790 }, { "grad_norm": 0.6016212701797485, "learning_rate": 9.474015815039446e-05, "loss": 0.0871, "step": 3800 }, { "grad_norm": 0.33667683601379395, "learning_rate": 9.470318653043565e-05, "loss": 0.0516, "step": 3810 }, { "grad_norm": 0.7373476028442383, "learning_rate": 9.466609269367185e-05, "loss": 0.0633, "step": 3820 }, { "grad_norm": 0.6312997937202454, "learning_rate": 9.46288767415162e-05, "loss": 0.0481, "step": 3830 }, { "grad_norm": 0.6646838188171387, "learning_rate": 9.459153877571567e-05, "loss": 0.0494, "step": 3840 }, { "grad_norm": 0.7411118745803833, "learning_rate": 9.455407889835087e-05, "loss": 0.0733, "step": 3850 }, { "grad_norm": 0.46237701177597046, "learning_rate": 9.451649721183564e-05, "loss": 0.0537, "step": 3860 }, { "grad_norm": 0.7067471146583557, "learning_rate": 9.447879381891692e-05, "loss": 0.0496, "step": 3870 }, { "grad_norm": 0.5780301690101624, "learning_rate": 9.444096882267428e-05, "loss": 0.0433, "step": 3880 }, { "grad_norm": 0.4460768699645996, "learning_rate": 9.440302232651988e-05, "loss": 0.0608, "step": 3890 }, { "grad_norm": 0.700219452381134, "learning_rate": 9.436495443419795e-05, "loss": 0.0526, "step": 3900 }, { "grad_norm": 0.5306059718132019, "learning_rate": 9.432676524978466e-05, "loss": 0.0501, "step": 3910 }, { "grad_norm": 0.741418182849884, "learning_rate": 9.42884548776878e-05, "loss": 0.0577, "step": 3920 }, { "grad_norm": 0.4041306674480438, "learning_rate": 9.425002342264646e-05, "loss": 0.0403, "step": 3930 }, { "grad_norm": 0.6558183431625366, "learning_rate": 9.421147098973077e-05, "loss": 0.0608, "step": 3940 }, { "grad_norm": 0.4879702925682068, "learning_rate": 9.41727976843416e-05, "loss": 0.0461, "step": 3950 }, { "grad_norm": 0.6670216917991638, "learning_rate": 9.413400361221029e-05, "loss": 0.0657, "step": 3960 }, { "grad_norm": 0.41735023260116577, "learning_rate": 9.409508887939835e-05, "loss": 0.0433, "step": 3970 }, { "grad_norm": 0.6116709113121033, "learning_rate": 9.40560535922972e-05, "loss": 0.054, "step": 3980 }, { "grad_norm": 0.586883544921875, "learning_rate": 9.40168978576278e-05, "loss": 0.0576, "step": 3990 }, { "grad_norm": 0.8781511187553406, "learning_rate": 9.397762178244043e-05, "loss": 0.0728, "step": 4000 }, { "grad_norm": 0.570734441280365, "learning_rate": 9.393822547411439e-05, "loss": 0.0482, "step": 4010 }, { "grad_norm": 0.2962997555732727, "learning_rate": 9.389870904035769e-05, "loss": 0.0514, "step": 4020 }, { "grad_norm": 0.7230541706085205, "learning_rate": 9.385907258920672e-05, "loss": 0.0599, "step": 4030 }, { "grad_norm": 0.43720558285713196, "learning_rate": 9.381931622902607e-05, "loss": 0.0433, "step": 4040 }, { "grad_norm": 0.40129533410072327, "learning_rate": 9.377944006850807e-05, "loss": 0.0654, "step": 4050 }, { "grad_norm": 0.4485349655151367, "learning_rate": 9.373944421667265e-05, "loss": 0.0443, "step": 4060 }, { "grad_norm": 0.5878545045852661, "learning_rate": 9.369932878286691e-05, "loss": 0.0387, "step": 4070 }, { "grad_norm": 0.8306893706321716, "learning_rate": 9.365909387676494e-05, "loss": 0.0687, "step": 4080 }, { "grad_norm": 0.5510880947113037, "learning_rate": 9.361873960836744e-05, "loss": 0.0514, "step": 4090 }, { "grad_norm": 0.5287096500396729, "learning_rate": 9.357826608800142e-05, "loss": 0.0457, "step": 4100 }, { "grad_norm": 0.7896179556846619, "learning_rate": 9.353767342631994e-05, "loss": 0.0586, "step": 4110 }, { "grad_norm": 0.612169086933136, "learning_rate": 9.34969617343018e-05, "loss": 0.0548, "step": 4120 }, { "grad_norm": 0.47709015011787415, "learning_rate": 9.345613112325122e-05, "loss": 0.0429, "step": 4130 }, { "grad_norm": 0.7442551851272583, "learning_rate": 9.34151817047975e-05, "loss": 0.0546, "step": 4140 }, { "grad_norm": 0.64883953332901, "learning_rate": 9.33741135908948e-05, "loss": 0.0661, "step": 4150 }, { "grad_norm": 0.3937470316886902, "learning_rate": 9.33329268938218e-05, "loss": 0.0476, "step": 4160 }, { "grad_norm": 0.8329939246177673, "learning_rate": 9.329162172618132e-05, "loss": 0.0529, "step": 4170 }, { "grad_norm": 0.47104158997535706, "learning_rate": 9.325019820090013e-05, "loss": 0.0716, "step": 4180 }, { "grad_norm": 0.7702105641365051, "learning_rate": 9.320865643122855e-05, "loss": 0.0638, "step": 4190 }, { "grad_norm": 0.4983150064945221, "learning_rate": 9.316699653074023e-05, "loss": 0.0538, "step": 4200 }, { "grad_norm": 0.5246564149856567, "learning_rate": 9.312521861333172e-05, "loss": 0.0554, "step": 4210 }, { "grad_norm": 0.2601536512374878, "learning_rate": 9.308332279322224e-05, "loss": 0.0522, "step": 4220 }, { "grad_norm": 0.6125774383544922, "learning_rate": 9.304130918495338e-05, "loss": 0.0662, "step": 4230 }, { "grad_norm": 0.7137991189956665, "learning_rate": 9.299917790338874e-05, "loss": 0.0694, "step": 4240 }, { "grad_norm": 0.4026471674442291, "learning_rate": 9.295692906371363e-05, "loss": 0.0738, "step": 4250 }, { "grad_norm": 0.6883469223976135, "learning_rate": 9.291456278143476e-05, "loss": 0.0507, "step": 4260 }, { "grad_norm": 0.7981429100036621, "learning_rate": 9.287207917237994e-05, "loss": 0.0505, "step": 4270 }, { "grad_norm": 0.6309172511100769, "learning_rate": 9.282947835269773e-05, "loss": 0.0405, "step": 4280 }, { "grad_norm": 0.7187209129333496, "learning_rate": 9.278676043885715e-05, "loss": 0.0511, "step": 4290 }, { "grad_norm": 0.6422325968742371, "learning_rate": 9.274392554764733e-05, "loss": 0.0495, "step": 4300 }, { "grad_norm": 0.28612279891967773, "learning_rate": 9.270097379617723e-05, "loss": 0.0428, "step": 4310 }, { "grad_norm": 0.4601300060749054, "learning_rate": 9.26579053018753e-05, "loss": 0.0849, "step": 4320 }, { "grad_norm": 0.22381387650966644, "learning_rate": 9.261472018248918e-05, "loss": 0.0487, "step": 4330 }, { "grad_norm": 0.43913787603378296, "learning_rate": 9.25714185560853e-05, "loss": 0.0491, "step": 4340 }, { "grad_norm": 0.46309322118759155, "learning_rate": 9.252800054104868e-05, "loss": 0.0473, "step": 4350 }, { "grad_norm": 0.6964513063430786, "learning_rate": 9.248446625608252e-05, "loss": 0.0584, "step": 4360 }, { "grad_norm": 0.6720502376556396, "learning_rate": 9.244081582020789e-05, "loss": 0.0563, "step": 4370 }, { "grad_norm": 0.7771483659744263, "learning_rate": 9.239704935276339e-05, "loss": 0.0586, "step": 4380 }, { "grad_norm": 0.6038665175437927, "learning_rate": 9.235316697340489e-05, "loss": 0.0509, "step": 4390 }, { "grad_norm": 0.5018665194511414, "learning_rate": 9.230916880210512e-05, "loss": 0.0464, "step": 4400 }, { "grad_norm": 0.5877880454063416, "learning_rate": 9.226505495915342e-05, "loss": 0.0445, "step": 4410 }, { "grad_norm": 0.5300883054733276, "learning_rate": 9.222082556515536e-05, "loss": 0.0663, "step": 4420 }, { "grad_norm": 0.5392895340919495, "learning_rate": 9.217648074103242e-05, "loss": 0.0559, "step": 4430 }, { "grad_norm": 0.5599974989891052, "learning_rate": 9.213202060802161e-05, "loss": 0.0537, "step": 4440 }, { "grad_norm": 0.6256566643714905, "learning_rate": 9.208744528767528e-05, "loss": 0.0635, "step": 4450 }, { "grad_norm": 0.3556930124759674, "learning_rate": 9.204275490186064e-05, "loss": 0.0505, "step": 4460 }, { "grad_norm": 0.6049244999885559, "learning_rate": 9.199794957275949e-05, "loss": 0.0394, "step": 4470 }, { "grad_norm": 0.35050565004348755, "learning_rate": 9.19530294228679e-05, "loss": 0.0564, "step": 4480 }, { "grad_norm": 0.6215412616729736, "learning_rate": 9.190799457499583e-05, "loss": 0.0545, "step": 4490 }, { "grad_norm": 0.6466807723045349, "learning_rate": 9.186284515226686e-05, "loss": 0.057, "step": 4500 }, { "grad_norm": 0.5352128148078918, "learning_rate": 9.181758127811777e-05, "loss": 0.0666, "step": 4510 }, { "grad_norm": 0.4230160713195801, "learning_rate": 9.177220307629825e-05, "loss": 0.0608, "step": 4520 }, { "grad_norm": 0.2667311429977417, "learning_rate": 9.172671067087059e-05, "loss": 0.0507, "step": 4530 }, { "grad_norm": 0.9667556881904602, "learning_rate": 9.16811041862093e-05, "loss": 0.0666, "step": 4540 }, { "grad_norm": 0.4570826590061188, "learning_rate": 9.163538374700076e-05, "loss": 0.0659, "step": 4550 }, { "grad_norm": 0.5891528129577637, "learning_rate": 9.158954947824287e-05, "loss": 0.0688, "step": 4560 }, { "grad_norm": 0.8018060326576233, "learning_rate": 9.154360150524482e-05, "loss": 0.0771, "step": 4570 }, { "grad_norm": 0.6970295310020447, "learning_rate": 9.14975399536266e-05, "loss": 0.0497, "step": 4580 }, { "grad_norm": 0.5423128008842468, "learning_rate": 9.14513649493187e-05, "loss": 0.0547, "step": 4590 }, { "grad_norm": 0.3745243549346924, "learning_rate": 9.140507661856187e-05, "loss": 0.0475, "step": 4600 }, { "grad_norm": 0.9668270945549011, "learning_rate": 9.135867508790661e-05, "loss": 0.0608, "step": 4610 }, { "grad_norm": 0.5553457140922546, "learning_rate": 9.131216048421291e-05, "loss": 0.0457, "step": 4620 }, { "grad_norm": 0.6711184978485107, "learning_rate": 9.126553293464998e-05, "loss": 0.062, "step": 4630 }, { "grad_norm": 0.723639965057373, "learning_rate": 9.121879256669572e-05, "loss": 0.0656, "step": 4640 }, { "grad_norm": 0.7574224472045898, "learning_rate": 9.117193950813652e-05, "loss": 0.0521, "step": 4650 }, { "grad_norm": 0.596269428730011, "learning_rate": 9.112497388706685e-05, "loss": 0.051, "step": 4660 }, { "grad_norm": 0.6881635785102844, "learning_rate": 9.10778958318889e-05, "loss": 0.0385, "step": 4670 }, { "grad_norm": 0.5309086441993713, "learning_rate": 9.103070547131232e-05, "loss": 0.0548, "step": 4680 }, { "grad_norm": 0.6249902248382568, "learning_rate": 9.098340293435375e-05, "loss": 0.0661, "step": 4690 }, { "grad_norm": 0.48949649930000305, "learning_rate": 9.093598835033649e-05, "loss": 0.0525, "step": 4700 }, { "grad_norm": 0.7062347531318665, "learning_rate": 9.088846184889021e-05, "loss": 0.0636, "step": 4710 }, { "grad_norm": 0.5817822813987732, "learning_rate": 9.084082355995057e-05, "loss": 0.0415, "step": 4720 }, { "grad_norm": 0.29488614201545715, "learning_rate": 9.079307361375882e-05, "loss": 0.0587, "step": 4730 }, { "grad_norm": 0.4921126961708069, "learning_rate": 9.074521214086149e-05, "loss": 0.051, "step": 4740 }, { "grad_norm": 0.5861870050430298, "learning_rate": 9.069723927211001e-05, "loss": 0.0483, "step": 4750 }, { "grad_norm": 0.5055922865867615, "learning_rate": 9.064915513866037e-05, "loss": 0.0572, "step": 4760 }, { "grad_norm": 0.8512992262840271, "learning_rate": 9.060095987197279e-05, "loss": 0.0516, "step": 4770 }, { "grad_norm": 0.6771171689033508, "learning_rate": 9.055265360381126e-05, "loss": 0.0793, "step": 4780 }, { "grad_norm": 0.4825901985168457, "learning_rate": 9.050423646624326e-05, "loss": 0.0497, "step": 4790 }, { "grad_norm": 0.4528404176235199, "learning_rate": 9.045570859163943e-05, "loss": 0.0522, "step": 4800 }, { "grad_norm": 0.5731555819511414, "learning_rate": 9.04070701126731e-05, "loss": 0.0554, "step": 4810 }, { "grad_norm": 0.2991088330745697, "learning_rate": 9.035832116232001e-05, "loss": 0.0559, "step": 4820 }, { "grad_norm": 0.3984524607658386, "learning_rate": 9.030946187385796e-05, "loss": 0.0744, "step": 4830 }, { "grad_norm": 0.4571893513202667, "learning_rate": 9.026049238086635e-05, "loss": 0.0563, "step": 4840 }, { "grad_norm": 0.6174187064170837, "learning_rate": 9.021141281722591e-05, "loss": 0.0459, "step": 4850 }, { "grad_norm": 0.4401225447654724, "learning_rate": 9.01622233171183e-05, "loss": 0.0573, "step": 4860 }, { "grad_norm": 0.3658362925052643, "learning_rate": 9.011292401502574e-05, "loss": 0.0545, "step": 4870 }, { "grad_norm": 0.6473553776741028, "learning_rate": 9.006351504573063e-05, "loss": 0.0445, "step": 4880 }, { "grad_norm": 0.47061866521835327, "learning_rate": 9.001399654431519e-05, "loss": 0.0558, "step": 4890 }, { "grad_norm": 0.5641303062438965, "learning_rate": 8.996436864616116e-05, "loss": 0.0399, "step": 4900 }, { "grad_norm": 0.5072057247161865, "learning_rate": 8.991463148694925e-05, "loss": 0.0453, "step": 4910 }, { "grad_norm": 0.34845972061157227, "learning_rate": 8.986478520265902e-05, "loss": 0.0658, "step": 4920 }, { "grad_norm": 0.3798791170120239, "learning_rate": 8.981482992956827e-05, "loss": 0.0487, "step": 4930 }, { "grad_norm": 0.5989367365837097, "learning_rate": 8.976476580425282e-05, "loss": 0.0535, "step": 4940 }, { "grad_norm": 0.5135841369628906, "learning_rate": 8.971459296358606e-05, "loss": 0.0483, "step": 4950 }, { "grad_norm": 0.4027482867240906, "learning_rate": 8.966431154473864e-05, "loss": 0.0466, "step": 4960 }, { "grad_norm": 0.3714529275894165, "learning_rate": 8.961392168517803e-05, "loss": 0.0579, "step": 4970 }, { "grad_norm": 0.597233235836029, "learning_rate": 8.956342352266821e-05, "loss": 0.0464, "step": 4980 }, { "grad_norm": 0.8222518563270569, "learning_rate": 8.95128171952692e-05, "loss": 0.0592, "step": 4990 }, { "grad_norm": 0.7420825362205505, "learning_rate": 8.946210284133676e-05, "loss": 0.0615, "step": 5000 }, { "grad_norm": 0.5680402517318726, "learning_rate": 8.941128059952201e-05, "loss": 0.0588, "step": 5010 }, { "grad_norm": 0.6809719800949097, "learning_rate": 8.936035060877102e-05, "loss": 0.063, "step": 5020 }, { "grad_norm": 0.4917536675930023, "learning_rate": 8.930931300832443e-05, "loss": 0.0418, "step": 5030 }, { "grad_norm": 0.6257069110870361, "learning_rate": 8.925816793771711e-05, "loss": 0.047, "step": 5040 }, { "grad_norm": 0.475649356842041, "learning_rate": 8.92069155367777e-05, "loss": 0.0489, "step": 5050 }, { "grad_norm": 0.3999846279621124, "learning_rate": 8.915555594562834e-05, "loss": 0.0325, "step": 5060 }, { "grad_norm": 0.49095281958580017, "learning_rate": 8.910408930468416e-05, "loss": 0.0603, "step": 5070 }, { "grad_norm": 0.45968499779701233, "learning_rate": 8.905251575465303e-05, "loss": 0.0658, "step": 5080 }, { "grad_norm": 0.555978536605835, "learning_rate": 8.900083543653502e-05, "loss": 0.0456, "step": 5090 }, { "grad_norm": 0.2715839445590973, "learning_rate": 8.894904849162218e-05, "loss": 0.0469, "step": 5100 }, { "grad_norm": 0.33360978960990906, "learning_rate": 8.889715506149802e-05, "loss": 0.05, "step": 5110 }, { "grad_norm": 0.5981850624084473, "learning_rate": 8.884515528803722e-05, "loss": 0.0625, "step": 5120 }, { "grad_norm": 0.3357233703136444, "learning_rate": 8.879304931340517e-05, "loss": 0.0581, "step": 5130 }, { "grad_norm": 0.3329426944255829, "learning_rate": 8.874083728005759e-05, "loss": 0.0524, "step": 5140 }, { "grad_norm": 0.40819597244262695, "learning_rate": 8.868851933074021e-05, "loss": 0.0417, "step": 5150 }, { "grad_norm": 0.4466436803340912, "learning_rate": 8.863609560848829e-05, "loss": 0.0537, "step": 5160 }, { "grad_norm": 0.6313127279281616, "learning_rate": 8.85835662566263e-05, "loss": 0.0513, "step": 5170 }, { "grad_norm": 0.4265729784965515, "learning_rate": 8.853093141876747e-05, "loss": 0.0401, "step": 5180 }, { "grad_norm": 0.5301429033279419, "learning_rate": 8.847819123881343e-05, "loss": 0.0676, "step": 5190 }, { "grad_norm": 0.5491750836372375, "learning_rate": 8.842534586095383e-05, "loss": 0.0419, "step": 5200 }, { "grad_norm": 0.49819865822792053, "learning_rate": 8.837239542966593e-05, "loss": 0.0515, "step": 5210 }, { "grad_norm": 0.4810800850391388, "learning_rate": 8.831934008971417e-05, "loss": 0.0464, "step": 5220 }, { "grad_norm": 0.47761476039886475, "learning_rate": 8.826617998614982e-05, "loss": 0.0353, "step": 5230 }, { "grad_norm": 0.5523068308830261, "learning_rate": 8.821291526431056e-05, "loss": 0.0439, "step": 5240 }, { "grad_norm": 0.43619635701179504, "learning_rate": 8.815954606982015e-05, "loss": 0.0636, "step": 5250 }, { "grad_norm": 1.0123933553695679, "learning_rate": 8.810607254858789e-05, "loss": 0.0503, "step": 5260 }, { "grad_norm": 0.4912901520729065, "learning_rate": 8.805249484680838e-05, "loss": 0.0415, "step": 5270 }, { "grad_norm": 0.3856203556060791, "learning_rate": 8.799881311096096e-05, "loss": 0.0549, "step": 5280 }, { "grad_norm": 0.39408963918685913, "learning_rate": 8.794502748780949e-05, "loss": 0.0574, "step": 5290 }, { "grad_norm": 0.28046944737434387, "learning_rate": 8.78911381244018e-05, "loss": 0.0451, "step": 5300 }, { "grad_norm": 0.4947446584701538, "learning_rate": 8.783714516806933e-05, "loss": 0.0428, "step": 5310 }, { "grad_norm": 0.41503649950027466, "learning_rate": 8.77830487664268e-05, "loss": 0.0479, "step": 5320 }, { "grad_norm": 0.5830925703048706, "learning_rate": 8.772884906737167e-05, "loss": 0.0637, "step": 5330 }, { "grad_norm": 0.6117151975631714, "learning_rate": 8.767454621908387e-05, "loss": 0.0552, "step": 5340 }, { "grad_norm": 0.36225998401641846, "learning_rate": 8.76201403700253e-05, "loss": 0.0393, "step": 5350 }, { "grad_norm": 0.33125925064086914, "learning_rate": 8.756563166893949e-05, "loss": 0.0485, "step": 5360 }, { "grad_norm": 0.36044541001319885, "learning_rate": 8.751102026485113e-05, "loss": 0.0434, "step": 5370 }, { "grad_norm": 0.4812445640563965, "learning_rate": 8.745630630706571e-05, "loss": 0.0322, "step": 5380 }, { "grad_norm": 0.2674048840999603, "learning_rate": 8.740148994516912e-05, "loss": 0.0394, "step": 5390 }, { "grad_norm": 0.5573847889900208, "learning_rate": 8.73465713290272e-05, "loss": 0.0438, "step": 5400 }, { "grad_norm": 0.3360580801963806, "learning_rate": 8.729155060878533e-05, "loss": 0.05, "step": 5410 }, { "grad_norm": 0.743016242980957, "learning_rate": 8.723642793486809e-05, "loss": 0.0465, "step": 5420 }, { "grad_norm": 0.5473203063011169, "learning_rate": 8.718120345797873e-05, "loss": 0.0432, "step": 5430 }, { "grad_norm": 0.7089893817901611, "learning_rate": 8.712587732909889e-05, "loss": 0.0711, "step": 5440 }, { "grad_norm": 0.6201307773590088, "learning_rate": 8.707044969948806e-05, "loss": 0.0398, "step": 5450 }, { "grad_norm": 0.4404928684234619, "learning_rate": 8.701492072068329e-05, "loss": 0.0307, "step": 5460 }, { "grad_norm": 0.18850098550319672, "learning_rate": 8.695929054449869e-05, "loss": 0.0389, "step": 5470 }, { "grad_norm": 0.46456393599510193, "learning_rate": 8.690355932302501e-05, "loss": 0.0524, "step": 5480 }, { "grad_norm": 0.42354366183280945, "learning_rate": 8.684772720862931e-05, "loss": 0.0457, "step": 5490 }, { "grad_norm": 0.5644344687461853, "learning_rate": 8.679179435395446e-05, "loss": 0.0449, "step": 5500 }, { "grad_norm": 0.4499945342540741, "learning_rate": 8.673576091191874e-05, "loss": 0.0581, "step": 5510 }, { "grad_norm": 0.3674764633178711, "learning_rate": 8.667962703571541e-05, "loss": 0.0543, "step": 5520 }, { "grad_norm": 0.4706595540046692, "learning_rate": 8.662339287881238e-05, "loss": 0.0568, "step": 5530 }, { "grad_norm": 0.26484641432762146, "learning_rate": 8.656705859495169e-05, "loss": 0.0442, "step": 5540 }, { "grad_norm": 0.3002794682979584, "learning_rate": 8.651062433814912e-05, "loss": 0.0471, "step": 5550 }, { "grad_norm": 0.4488489627838135, "learning_rate": 8.645409026269375e-05, "loss": 0.0434, "step": 5560 }, { "grad_norm": 0.5279734134674072, "learning_rate": 8.639745652314759e-05, "loss": 0.0537, "step": 5570 }, { "grad_norm": 0.3375439941883087, "learning_rate": 8.634072327434515e-05, "loss": 0.0461, "step": 5580 }, { "grad_norm": 0.4130055904388428, "learning_rate": 8.628389067139294e-05, "loss": 0.0366, "step": 5590 }, { "grad_norm": 0.28176242113113403, "learning_rate": 8.622695886966911e-05, "loss": 0.0601, "step": 5600 }, { "grad_norm": 0.3169611692428589, "learning_rate": 8.616992802482308e-05, "loss": 0.0432, "step": 5610 }, { "grad_norm": 0.25434207916259766, "learning_rate": 8.611279829277496e-05, "loss": 0.0402, "step": 5620 }, { "grad_norm": 0.36730343103408813, "learning_rate": 8.605556982971528e-05, "loss": 0.0317, "step": 5630 }, { "grad_norm": 0.38030752539634705, "learning_rate": 8.599824279210447e-05, "loss": 0.048, "step": 5640 }, { "grad_norm": 0.27405062317848206, "learning_rate": 8.594081733667243e-05, "loss": 0.053, "step": 5650 }, { "grad_norm": 0.3755406141281128, "learning_rate": 8.58832936204182e-05, "loss": 0.0369, "step": 5660 }, { "grad_norm": 0.5103726983070374, "learning_rate": 8.582567180060942e-05, "loss": 0.0421, "step": 5670 }, { "grad_norm": 0.47422799468040466, "learning_rate": 8.576795203478194e-05, "loss": 0.0474, "step": 5680 }, { "grad_norm": 0.4571792185306549, "learning_rate": 8.571013448073939e-05, "loss": 0.0437, "step": 5690 }, { "grad_norm": 0.2985520660877228, "learning_rate": 8.565221929655275e-05, "loss": 0.0283, "step": 5700 }, { "grad_norm": 0.33401042222976685, "learning_rate": 8.559420664055992e-05, "loss": 0.0485, "step": 5710 }, { "grad_norm": 0.4235278367996216, "learning_rate": 8.553609667136532e-05, "loss": 0.0335, "step": 5720 }, { "grad_norm": 0.44519418478012085, "learning_rate": 8.547788954783936e-05, "loss": 0.0424, "step": 5730 }, { "grad_norm": 0.44941434264183044, "learning_rate": 8.541958542911808e-05, "loss": 0.0449, "step": 5740 }, { "grad_norm": 0.30726078152656555, "learning_rate": 8.536118447460275e-05, "loss": 0.0349, "step": 5750 }, { "grad_norm": 0.39587411284446716, "learning_rate": 8.530268684395932e-05, "loss": 0.0414, "step": 5760 }, { "grad_norm": 0.464193195104599, "learning_rate": 8.524409269711807e-05, "loss": 0.0448, "step": 5770 }, { "grad_norm": 0.27170541882514954, "learning_rate": 8.51854021942732e-05, "loss": 0.0528, "step": 5780 }, { "grad_norm": 0.5538057684898376, "learning_rate": 8.512661549588227e-05, "loss": 0.0453, "step": 5790 }, { "grad_norm": 0.4806496202945709, "learning_rate": 8.506773276266588e-05, "loss": 0.0494, "step": 5800 }, { "grad_norm": 0.45044761896133423, "learning_rate": 8.500875415560721e-05, "loss": 0.0331, "step": 5810 }, { "grad_norm": 0.23951706290245056, "learning_rate": 8.494967983595144e-05, "loss": 0.0378, "step": 5820 }, { "grad_norm": 0.5738400220870972, "learning_rate": 8.489050996520558e-05, "loss": 0.051, "step": 5830 }, { "grad_norm": 0.3326394855976105, "learning_rate": 8.483124470513775e-05, "loss": 0.0417, "step": 5840 }, { "grad_norm": 0.34265029430389404, "learning_rate": 8.477188421777692e-05, "loss": 0.0325, "step": 5850 }, { "grad_norm": 0.29334756731987, "learning_rate": 8.47124286654124e-05, "loss": 0.0414, "step": 5860 }, { "grad_norm": 0.39102500677108765, "learning_rate": 8.465287821059341e-05, "loss": 0.0445, "step": 5870 }, { "grad_norm": 0.45586511492729187, "learning_rate": 8.45932330161286e-05, "loss": 0.045, "step": 5880 }, { "grad_norm": 0.3896578848361969, "learning_rate": 8.453349324508567e-05, "loss": 0.0461, "step": 5890 }, { "grad_norm": 0.7093162536621094, "learning_rate": 8.447365906079088e-05, "loss": 0.0616, "step": 5900 }, { "grad_norm": 0.35690462589263916, "learning_rate": 8.441373062682856e-05, "loss": 0.0559, "step": 5910 }, { "grad_norm": 0.5989769697189331, "learning_rate": 8.43537081070408e-05, "loss": 0.043, "step": 5920 }, { "grad_norm": 0.5885160565376282, "learning_rate": 8.429359166552689e-05, "loss": 0.0383, "step": 5930 }, { "grad_norm": 0.25422751903533936, "learning_rate": 8.423338146664284e-05, "loss": 0.0506, "step": 5940 }, { "grad_norm": 0.5703669190406799, "learning_rate": 8.417307767500107e-05, "loss": 0.0435, "step": 5950 }, { "grad_norm": 0.2736944854259491, "learning_rate": 8.411268045546983e-05, "loss": 0.0415, "step": 5960 }, { "grad_norm": 1.0298906564712524, "learning_rate": 8.405218997317281e-05, "loss": 0.0557, "step": 5970 }, { "grad_norm": 0.7170928120613098, "learning_rate": 8.399160639348869e-05, "loss": 0.0486, "step": 5980 }, { "grad_norm": 0.7026037573814392, "learning_rate": 8.393092988205065e-05, "loss": 0.0585, "step": 5990 }, { "grad_norm": 0.3382510840892792, "learning_rate": 8.387016060474597e-05, "loss": 0.0652, "step": 6000 }, { "grad_norm": 0.4429570734500885, "learning_rate": 8.380929872771551e-05, "loss": 0.0462, "step": 6010 }, { "grad_norm": 0.2875368297100067, "learning_rate": 8.374834441735335e-05, "loss": 0.0514, "step": 6020 }, { "grad_norm": 0.36912545561790466, "learning_rate": 8.368729784030622e-05, "loss": 0.0462, "step": 6030 }, { "grad_norm": 0.36750081181526184, "learning_rate": 8.362615916347315e-05, "loss": 0.036, "step": 6040 }, { "grad_norm": 0.47908490896224976, "learning_rate": 8.356492855400493e-05, "loss": 0.05, "step": 6050 }, { "grad_norm": 0.44548559188842773, "learning_rate": 8.350360617930371e-05, "loss": 0.0411, "step": 6060 }, { "grad_norm": 0.7380837798118591, "learning_rate": 8.344219220702255e-05, "loss": 0.0534, "step": 6070 }, { "grad_norm": 0.6426033973693848, "learning_rate": 8.338068680506485e-05, "loss": 0.0561, "step": 6080 }, { "grad_norm": 0.47584810853004456, "learning_rate": 8.33190901415841e-05, "loss": 0.0445, "step": 6090 }, { "grad_norm": 0.588601291179657, "learning_rate": 8.325740238498317e-05, "loss": 0.0471, "step": 6100 }, { "grad_norm": 0.6466059684753418, "learning_rate": 8.319562370391406e-05, "loss": 0.0342, "step": 6110 }, { "grad_norm": 0.48216512799263, "learning_rate": 8.31337542672773e-05, "loss": 0.0403, "step": 6120 }, { "grad_norm": 0.5129768252372742, "learning_rate": 8.307179424422158e-05, "loss": 0.0372, "step": 6130 }, { "grad_norm": 0.2967594265937805, "learning_rate": 8.300974380414327e-05, "loss": 0.0456, "step": 6140 }, { "grad_norm": 0.6408554911613464, "learning_rate": 8.294760311668586e-05, "loss": 0.062, "step": 6150 }, { "grad_norm": 0.6735295057296753, "learning_rate": 8.288537235173961e-05, "loss": 0.0707, "step": 6160 }, { "grad_norm": 0.5444054007530212, "learning_rate": 8.282305167944108e-05, "loss": 0.0575, "step": 6170 }, { "grad_norm": 0.7867200374603271, "learning_rate": 8.276064127017262e-05, "loss": 0.0518, "step": 6180 }, { "grad_norm": 0.3864005208015442, "learning_rate": 8.269814129456189e-05, "loss": 0.0638, "step": 6190 }, { "grad_norm": 0.38642510771751404, "learning_rate": 8.263555192348143e-05, "loss": 0.0416, "step": 6200 }, { "grad_norm": 0.414685994386673, "learning_rate": 8.257287332804819e-05, "loss": 0.0482, "step": 6210 }, { "grad_norm": 0.4706995487213135, "learning_rate": 8.251010567962307e-05, "loss": 0.0394, "step": 6220 }, { "grad_norm": 0.3204301595687866, "learning_rate": 8.244724914981041e-05, "loss": 0.0415, "step": 6230 }, { "grad_norm": 0.3907712697982788, "learning_rate": 8.238430391045757e-05, "loss": 0.0383, "step": 6240 }, { "grad_norm": 0.43203648924827576, "learning_rate": 8.232127013365445e-05, "loss": 0.0498, "step": 6250 }, { "grad_norm": 0.45337560772895813, "learning_rate": 8.225814799173295e-05, "loss": 0.0413, "step": 6260 }, { "grad_norm": 0.31500765681266785, "learning_rate": 8.219493765726663e-05, "loss": 0.0412, "step": 6270 }, { "grad_norm": 0.6452749371528625, "learning_rate": 8.21316393030701e-05, "loss": 0.0567, "step": 6280 }, { "grad_norm": 0.3556128740310669, "learning_rate": 8.206825310219865e-05, "loss": 0.0418, "step": 6290 }, { "grad_norm": 0.3755437135696411, "learning_rate": 8.200477922794776e-05, "loss": 0.0586, "step": 6300 }, { "grad_norm": 0.3259917199611664, "learning_rate": 8.194121785385256e-05, "loss": 0.039, "step": 6310 }, { "grad_norm": 0.5897827744483948, "learning_rate": 8.187756915368741e-05, "loss": 0.0397, "step": 6320 }, { "grad_norm": 0.4412654638290405, "learning_rate": 8.181383330146544e-05, "loss": 0.0394, "step": 6330 }, { "grad_norm": 0.6785094141960144, "learning_rate": 8.175001047143804e-05, "loss": 0.0647, "step": 6340 }, { "grad_norm": 0.4786268174648285, "learning_rate": 8.168610083809438e-05, "loss": 0.0372, "step": 6350 }, { "grad_norm": 0.4114820957183838, "learning_rate": 8.162210457616095e-05, "loss": 0.0405, "step": 6360 }, { "grad_norm": 0.5983986258506775, "learning_rate": 8.155802186060109e-05, "loss": 0.0404, "step": 6370 }, { "grad_norm": 0.34187790751457214, "learning_rate": 8.149385286661453e-05, "loss": 0.0384, "step": 6380 }, { "grad_norm": 0.6770147681236267, "learning_rate": 8.14295977696368e-05, "loss": 0.0438, "step": 6390 }, { "grad_norm": 1.088093876838684, "learning_rate": 8.13652567453389e-05, "loss": 0.0515, "step": 6400 }, { "grad_norm": 0.3024458885192871, "learning_rate": 8.130082996962676e-05, "loss": 0.0369, "step": 6410 }, { "grad_norm": 0.31531649827957153, "learning_rate": 8.123631761864068e-05, "loss": 0.0458, "step": 6420 }, { "grad_norm": 0.4136103093624115, "learning_rate": 8.1171719868755e-05, "loss": 0.0421, "step": 6430 }, { "grad_norm": 0.41486772894859314, "learning_rate": 8.110703689657748e-05, "loss": 0.0374, "step": 6440 }, { "grad_norm": 0.5491014719009399, "learning_rate": 8.104226887894892e-05, "loss": 0.0381, "step": 6450 }, { "grad_norm": 0.6629388928413391, "learning_rate": 8.097741599294257e-05, "loss": 0.037, "step": 6460 }, { "grad_norm": 0.42994698882102966, "learning_rate": 8.091247841586378e-05, "loss": 0.0398, "step": 6470 }, { "grad_norm": 0.38615524768829346, "learning_rate": 8.084745632524939e-05, "loss": 0.0329, "step": 6480 }, { "grad_norm": 0.52663254737854, "learning_rate": 8.07823498988673e-05, "loss": 0.0459, "step": 6490 }, { "grad_norm": 0.3121413588523865, "learning_rate": 8.071715931471602e-05, "loss": 0.0316, "step": 6500 }, { "grad_norm": 0.2786811292171478, "learning_rate": 8.06518847510241e-05, "loss": 0.0305, "step": 6510 }, { "grad_norm": 0.3595089316368103, "learning_rate": 8.058652638624971e-05, "loss": 0.0347, "step": 6520 }, { "grad_norm": 0.3069809675216675, "learning_rate": 8.052108439908013e-05, "loss": 0.0342, "step": 6530 }, { "grad_norm": 0.5630214810371399, "learning_rate": 8.045555896843125e-05, "loss": 0.0481, "step": 6540 }, { "grad_norm": 0.5187119245529175, "learning_rate": 8.03899502734471e-05, "loss": 0.0352, "step": 6550 }, { "grad_norm": 0.5507297515869141, "learning_rate": 8.032425849349931e-05, "loss": 0.0526, "step": 6560 }, { "grad_norm": 0.30713117122650146, "learning_rate": 8.025848380818674e-05, "loss": 0.0367, "step": 6570 }, { "grad_norm": 0.38138148188591003, "learning_rate": 8.019262639733487e-05, "loss": 0.0406, "step": 6580 }, { "grad_norm": 0.8942192792892456, "learning_rate": 8.012668644099531e-05, "loss": 0.0475, "step": 6590 }, { "grad_norm": 0.22848564386367798, "learning_rate": 8.006066411944542e-05, "loss": 0.0396, "step": 6600 }, { "grad_norm": 0.5707852840423584, "learning_rate": 7.999455961318769e-05, "loss": 0.0335, "step": 6610 }, { "grad_norm": 0.5564255118370056, "learning_rate": 7.992837310294932e-05, "loss": 0.0466, "step": 6620 }, { "grad_norm": 0.6214336156845093, "learning_rate": 7.986210476968167e-05, "loss": 0.0529, "step": 6630 }, { "grad_norm": 0.3645685017108917, "learning_rate": 7.97957547945599e-05, "loss": 0.0421, "step": 6640 }, { "grad_norm": 0.41240009665489197, "learning_rate": 7.972932335898226e-05, "loss": 0.0611, "step": 6650 }, { "grad_norm": 0.38685816526412964, "learning_rate": 7.966281064456975e-05, "loss": 0.0318, "step": 6660 }, { "grad_norm": 0.6569758057594299, "learning_rate": 7.959621683316563e-05, "loss": 0.0481, "step": 6670 }, { "grad_norm": 0.34972772002220154, "learning_rate": 7.952954210683481e-05, "loss": 0.0431, "step": 6680 }, { "grad_norm": 0.31339651346206665, "learning_rate": 7.946278664786345e-05, "loss": 0.0402, "step": 6690 }, { "grad_norm": 0.465017706155777, "learning_rate": 7.939595063875842e-05, "loss": 0.0417, "step": 6700 }, { "grad_norm": 0.4094502925872803, "learning_rate": 7.932903426224683e-05, "loss": 0.0457, "step": 6710 }, { "grad_norm": 0.4140152335166931, "learning_rate": 7.926203770127552e-05, "loss": 0.0478, "step": 6720 }, { "grad_norm": 0.2585093379020691, "learning_rate": 7.919496113901046e-05, "loss": 0.0413, "step": 6730 }, { "grad_norm": 0.3717402517795563, "learning_rate": 7.912780475883649e-05, "loss": 0.0608, "step": 6740 }, { "grad_norm": 0.3836466372013092, "learning_rate": 7.906056874435652e-05, "loss": 0.0503, "step": 6750 }, { "grad_norm": 0.390178382396698, "learning_rate": 7.899325327939131e-05, "loss": 0.0425, "step": 6760 }, { "grad_norm": 0.39616844058036804, "learning_rate": 7.892585854797872e-05, "loss": 0.0457, "step": 6770 }, { "grad_norm": 0.5933540463447571, "learning_rate": 7.88583847343734e-05, "loss": 0.0416, "step": 6780 }, { "grad_norm": 0.47818222641944885, "learning_rate": 7.879083202304616e-05, "loss": 0.0323, "step": 6790 }, { "grad_norm": 0.5178842544555664, "learning_rate": 7.872320059868355e-05, "loss": 0.0418, "step": 6800 }, { "grad_norm": 0.43127959966659546, "learning_rate": 7.865549064618729e-05, "loss": 0.039, "step": 6810 }, { "grad_norm": 0.4189041256904602, "learning_rate": 7.858770235067381e-05, "loss": 0.0432, "step": 6820 }, { "grad_norm": 0.5595656037330627, "learning_rate": 7.851983589747374e-05, "loss": 0.0375, "step": 6830 }, { "grad_norm": 0.5023381114006042, "learning_rate": 7.845189147213133e-05, "loss": 0.0522, "step": 6840 }, { "grad_norm": 0.48145803809165955, "learning_rate": 7.838386926040407e-05, "loss": 0.047, "step": 6850 }, { "grad_norm": 0.3289065957069397, "learning_rate": 7.83157694482621e-05, "loss": 0.028, "step": 6860 }, { "grad_norm": 0.39679503440856934, "learning_rate": 7.824759222188768e-05, "loss": 0.0547, "step": 6870 }, { "grad_norm": 0.2578141391277313, "learning_rate": 7.817933776767478e-05, "loss": 0.0516, "step": 6880 }, { "grad_norm": 0.6736020445823669, "learning_rate": 7.811100627222842e-05, "loss": 0.0484, "step": 6890 }, { "grad_norm": 0.4434877038002014, "learning_rate": 7.804259792236435e-05, "loss": 0.0559, "step": 6900 }, { "grad_norm": 0.6563153266906738, "learning_rate": 7.797411290510835e-05, "loss": 0.0452, "step": 6910 }, { "grad_norm": 0.6429451704025269, "learning_rate": 7.790555140769586e-05, "loss": 0.0569, "step": 6920 }, { "grad_norm": 0.5441063046455383, "learning_rate": 7.78369136175714e-05, "loss": 0.0512, "step": 6930 }, { "grad_norm": 0.31242433190345764, "learning_rate": 7.776819972238806e-05, "loss": 0.0377, "step": 6940 }, { "grad_norm": 0.26298272609710693, "learning_rate": 7.7699409910007e-05, "loss": 0.0356, "step": 6950 }, { "grad_norm": 0.35689061880111694, "learning_rate": 7.763054436849694e-05, "loss": 0.0369, "step": 6960 }, { "grad_norm": 0.4908151924610138, "learning_rate": 7.756160328613364e-05, "loss": 0.043, "step": 6970 }, { "grad_norm": 0.3438277542591095, "learning_rate": 7.749258685139942e-05, "loss": 0.0464, "step": 6980 }, { "grad_norm": 0.468595951795578, "learning_rate": 7.742349525298253e-05, "loss": 0.0402, "step": 6990 }, { "grad_norm": 0.36681365966796875, "learning_rate": 7.735432867977679e-05, "loss": 0.041, "step": 7000 }, { "grad_norm": 0.3225845694541931, "learning_rate": 7.728508732088096e-05, "loss": 0.0498, "step": 7010 }, { "grad_norm": 0.6380502581596375, "learning_rate": 7.721577136559825e-05, "loss": 0.0432, "step": 7020 }, { "grad_norm": 0.4817630648612976, "learning_rate": 7.714638100343588e-05, "loss": 0.0621, "step": 7030 }, { "grad_norm": 0.6248610615730286, "learning_rate": 7.707691642410444e-05, "loss": 0.0641, "step": 7040 }, { "grad_norm": 0.5478596687316895, "learning_rate": 7.70073778175174e-05, "loss": 0.0346, "step": 7050 }, { "grad_norm": 0.398355633020401, "learning_rate": 7.69377653737907e-05, "loss": 0.0479, "step": 7060 }, { "grad_norm": 0.4127849042415619, "learning_rate": 7.686807928324209e-05, "loss": 0.0514, "step": 7070 }, { "grad_norm": 0.3307373523712158, "learning_rate": 7.679831973639065e-05, "loss": 0.0452, "step": 7080 }, { "grad_norm": 0.6282635927200317, "learning_rate": 7.672848692395637e-05, "loss": 0.0455, "step": 7090 }, { "grad_norm": 0.5247623324394226, "learning_rate": 7.665858103685944e-05, "loss": 0.0364, "step": 7100 }, { "grad_norm": 0.44411900639533997, "learning_rate": 7.658860226621991e-05, "loss": 0.042, "step": 7110 }, { "grad_norm": 0.22760289907455444, "learning_rate": 7.651855080335708e-05, "loss": 0.0301, "step": 7120 }, { "grad_norm": 0.2277648150920868, "learning_rate": 7.644842683978896e-05, "loss": 0.0399, "step": 7130 }, { "grad_norm": 0.5388209819793701, "learning_rate": 7.63782305672318e-05, "loss": 0.0502, "step": 7140 }, { "grad_norm": 0.20562751591205597, "learning_rate": 7.63079621775995e-05, "loss": 0.0435, "step": 7150 }, { "grad_norm": 0.791567862033844, "learning_rate": 7.623762186300319e-05, "loss": 0.0573, "step": 7160 }, { "grad_norm": 0.30695784091949463, "learning_rate": 7.616720981575057e-05, "loss": 0.0344, "step": 7170 }, { "grad_norm": 0.5200401544570923, "learning_rate": 7.609672622834552e-05, "loss": 0.0428, "step": 7180 }, { "grad_norm": 0.2493307888507843, "learning_rate": 7.602617129348747e-05, "loss": 0.0391, "step": 7190 }, { "grad_norm": 0.42052096128463745, "learning_rate": 7.595554520407088e-05, "loss": 0.0322, "step": 7200 }, { "grad_norm": 0.33217257261276245, "learning_rate": 7.588484815318484e-05, "loss": 0.0424, "step": 7210 }, { "grad_norm": 0.5206419825553894, "learning_rate": 7.581408033411234e-05, "loss": 0.0403, "step": 7220 }, { "grad_norm": 0.3290882706642151, "learning_rate": 7.574324194032995e-05, "loss": 0.0404, "step": 7230 }, { "grad_norm": 0.4738713204860687, "learning_rate": 7.567233316550705e-05, "loss": 0.0514, "step": 7240 }, { "grad_norm": 0.33865848183631897, "learning_rate": 7.560135420350562e-05, "loss": 0.0462, "step": 7250 }, { "grad_norm": 0.18623845279216766, "learning_rate": 7.553030524837935e-05, "loss": 0.0247, "step": 7260 }, { "grad_norm": 0.1978921741247177, "learning_rate": 7.545918649437341e-05, "loss": 0.0324, "step": 7270 }, { "grad_norm": 0.5654398798942566, "learning_rate": 7.538799813592377e-05, "loss": 0.041, "step": 7280 }, { "grad_norm": 0.48351162672042847, "learning_rate": 7.531674036765662e-05, "loss": 0.0419, "step": 7290 }, { "grad_norm": 0.2131793200969696, "learning_rate": 7.524541338438807e-05, "loss": 0.031, "step": 7300 }, { "grad_norm": 0.37112846970558167, "learning_rate": 7.517401738112328e-05, "loss": 0.0448, "step": 7310 }, { "grad_norm": 0.3980337977409363, "learning_rate": 7.510255255305628e-05, "loss": 0.0299, "step": 7320 }, { "grad_norm": 0.39956146478652954, "learning_rate": 7.503101909556911e-05, "loss": 0.0445, "step": 7330 }, { "grad_norm": 0.3776455521583557, "learning_rate": 7.495941720423154e-05, "loss": 0.0393, "step": 7340 }, { "grad_norm": 0.4671146273612976, "learning_rate": 7.488774707480042e-05, "loss": 0.0281, "step": 7350 }, { "grad_norm": 0.31814056634902954, "learning_rate": 7.481600890321911e-05, "loss": 0.0363, "step": 7360 }, { "grad_norm": 0.48548802733421326, "learning_rate": 7.474420288561708e-05, "loss": 0.0404, "step": 7370 }, { "grad_norm": 0.39891475439071655, "learning_rate": 7.467232921830921e-05, "loss": 0.0425, "step": 7380 }, { "grad_norm": 0.6269211769104004, "learning_rate": 7.460038809779537e-05, "loss": 0.0502, "step": 7390 }, { "grad_norm": 0.33319777250289917, "learning_rate": 7.452837972075983e-05, "loss": 0.0359, "step": 7400 }, { "grad_norm": 0.3188859522342682, "learning_rate": 7.445630428407074e-05, "loss": 0.0369, "step": 7410 }, { "grad_norm": 0.4725710153579712, "learning_rate": 7.43841619847796e-05, "loss": 0.0476, "step": 7420 }, { "grad_norm": 0.39266884326934814, "learning_rate": 7.431195302012072e-05, "loss": 0.0473, "step": 7430 }, { "grad_norm": 0.36300551891326904, "learning_rate": 7.423967758751061e-05, "loss": 0.0373, "step": 7440 }, { "grad_norm": 0.29192590713500977, "learning_rate": 7.416733588454758e-05, "loss": 0.0314, "step": 7450 }, { "grad_norm": 0.3165670335292816, "learning_rate": 7.409492810901106e-05, "loss": 0.0364, "step": 7460 }, { "grad_norm": 0.5099138617515564, "learning_rate": 7.402245445886116e-05, "loss": 0.0449, "step": 7470 }, { "grad_norm": 0.6238443851470947, "learning_rate": 7.394991513223806e-05, "loss": 0.0494, "step": 7480 }, { "grad_norm": 0.6527758240699768, "learning_rate": 7.38773103274615e-05, "loss": 0.0405, "step": 7490 }, { "grad_norm": 0.3557003140449524, "learning_rate": 7.380464024303028e-05, "loss": 0.0344, "step": 7500 }, { "grad_norm": 0.27560967206954956, "learning_rate": 7.373190507762162e-05, "loss": 0.0398, "step": 7510 }, { "grad_norm": 0.2316887378692627, "learning_rate": 7.365910503009066e-05, "loss": 0.0326, "step": 7520 }, { "grad_norm": 0.785270094871521, "learning_rate": 7.358624029946996e-05, "loss": 0.0329, "step": 7530 }, { "grad_norm": 0.2258644700050354, "learning_rate": 7.351331108496893e-05, "loss": 0.043, "step": 7540 }, { "grad_norm": 0.4553779661655426, "learning_rate": 7.344031758597325e-05, "loss": 0.035, "step": 7550 }, { "grad_norm": 0.32569703459739685, "learning_rate": 7.336726000204435e-05, "loss": 0.0403, "step": 7560 }, { "grad_norm": 0.4431248605251312, "learning_rate": 7.32941385329189e-05, "loss": 0.0382, "step": 7570 }, { "grad_norm": 0.5144472122192383, "learning_rate": 7.322095337850816e-05, "loss": 0.0481, "step": 7580 }, { "grad_norm": 0.44784116744995117, "learning_rate": 7.314770473889758e-05, "loss": 0.0449, "step": 7590 }, { "grad_norm": 0.5323681831359863, "learning_rate": 7.307439281434615e-05, "loss": 0.0508, "step": 7600 }, { "grad_norm": 0.21168549358844757, "learning_rate": 7.300101780528585e-05, "loss": 0.04, "step": 7610 }, { "grad_norm": 0.42234864830970764, "learning_rate": 7.292757991232117e-05, "loss": 0.0351, "step": 7620 }, { "grad_norm": 0.2938712239265442, "learning_rate": 7.285407933622848e-05, "loss": 0.0437, "step": 7630 }, { "grad_norm": 0.4372832477092743, "learning_rate": 7.278051627795557e-05, "loss": 0.0365, "step": 7640 }, { "grad_norm": 0.2331153005361557, "learning_rate": 7.270689093862105e-05, "loss": 0.0344, "step": 7650 }, { "grad_norm": 0.4866470992565155, "learning_rate": 7.263320351951374e-05, "loss": 0.0368, "step": 7660 }, { "grad_norm": 0.48618564009666443, "learning_rate": 7.255945422209227e-05, "loss": 0.0402, "step": 7670 }, { "grad_norm": 0.5336203575134277, "learning_rate": 7.248564324798437e-05, "loss": 0.0412, "step": 7680 }, { "grad_norm": 0.39618951082229614, "learning_rate": 7.241177079898644e-05, "loss": 0.0417, "step": 7690 }, { "grad_norm": 0.37101486325263977, "learning_rate": 7.233783707706295e-05, "loss": 0.0308, "step": 7700 }, { "grad_norm": 0.46168020367622375, "learning_rate": 7.226384228434586e-05, "loss": 0.045, "step": 7710 }, { "grad_norm": 0.5402097702026367, "learning_rate": 7.21897866231341e-05, "loss": 0.0402, "step": 7720 }, { "grad_norm": 0.22946733236312866, "learning_rate": 7.211567029589303e-05, "loss": 0.0449, "step": 7730 }, { "grad_norm": 0.34321534633636475, "learning_rate": 7.204149350525387e-05, "loss": 0.0402, "step": 7740 }, { "grad_norm": 0.27290424704551697, "learning_rate": 7.196725645401309e-05, "loss": 0.0333, "step": 7750 }, { "grad_norm": 0.5110877156257629, "learning_rate": 7.1892959345132e-05, "loss": 0.0323, "step": 7760 }, { "grad_norm": 0.2606765031814575, "learning_rate": 7.181860238173605e-05, "loss": 0.0489, "step": 7770 }, { "grad_norm": 0.5061051845550537, "learning_rate": 7.174418576711432e-05, "loss": 0.0445, "step": 7780 }, { "grad_norm": 0.5962272882461548, "learning_rate": 7.1669709704719e-05, "loss": 0.0539, "step": 7790 }, { "grad_norm": 0.4701775312423706, "learning_rate": 7.159517439816481e-05, "loss": 0.0488, "step": 7800 }, { "grad_norm": 0.2271687090396881, "learning_rate": 7.152058005122842e-05, "loss": 0.0305, "step": 7810 }, { "grad_norm": 0.24474292993545532, "learning_rate": 7.144592686784793e-05, "loss": 0.0232, "step": 7820 }, { "grad_norm": 0.3326061964035034, "learning_rate": 7.137121505212229e-05, "loss": 0.0431, "step": 7830 }, { "grad_norm": 0.28043511509895325, "learning_rate": 7.129644480831077e-05, "loss": 0.031, "step": 7840 }, { "grad_norm": 0.30489447712898254, "learning_rate": 7.122161634083234e-05, "loss": 0.0325, "step": 7850 }, { "grad_norm": 0.4969082474708557, "learning_rate": 7.114672985426516e-05, "loss": 0.0432, "step": 7860 }, { "grad_norm": 0.38436904549598694, "learning_rate": 7.107178555334606e-05, "loss": 0.0392, "step": 7870 }, { "grad_norm": 0.45061829686164856, "learning_rate": 7.099678364296989e-05, "loss": 0.0457, "step": 7880 }, { "grad_norm": 0.33805397152900696, "learning_rate": 7.0921724328189e-05, "loss": 0.0394, "step": 7890 }, { "grad_norm": 0.34628984332084656, "learning_rate": 7.084660781421268e-05, "loss": 0.0488, "step": 7900 }, { "grad_norm": 0.2685685157775879, "learning_rate": 7.077143430640662e-05, "loss": 0.0336, "step": 7910 }, { "grad_norm": 0.4601621925830841, "learning_rate": 7.069620401029232e-05, "loss": 0.0606, "step": 7920 }, { "grad_norm": 0.353318452835083, "learning_rate": 7.062091713154655e-05, "loss": 0.0307, "step": 7930 }, { "grad_norm": 0.36406975984573364, "learning_rate": 7.054557387600075e-05, "loss": 0.0369, "step": 7940 }, { "grad_norm": 0.48223838210105896, "learning_rate": 7.04701744496405e-05, "loss": 0.0431, "step": 7950 }, { "grad_norm": 0.2531633675098419, "learning_rate": 7.039471905860495e-05, "loss": 0.0333, "step": 7960 }, { "grad_norm": 0.3868260681629181, "learning_rate": 7.031920790918628e-05, "loss": 0.04, "step": 7970 }, { "grad_norm": 0.49379217624664307, "learning_rate": 7.024364120782906e-05, "loss": 0.0393, "step": 7980 }, { "grad_norm": 0.1904933750629425, "learning_rate": 7.016801916112978e-05, "loss": 0.034, "step": 7990 }, { "grad_norm": 0.2884490489959717, "learning_rate": 7.009234197583623e-05, "loss": 0.0362, "step": 8000 }, { "grad_norm": 0.4237913489341736, "learning_rate": 7.001660985884692e-05, "loss": 0.0398, "step": 8010 }, { "grad_norm": 0.23363447189331055, "learning_rate": 6.994082301721063e-05, "loss": 0.0412, "step": 8020 }, { "grad_norm": 0.4389555752277374, "learning_rate": 6.986498165812563e-05, "loss": 0.033, "step": 8030 }, { "grad_norm": 0.35672852396965027, "learning_rate": 6.978908598893932e-05, "loss": 0.0276, "step": 8040 }, { "grad_norm": 0.5513891577720642, "learning_rate": 6.971313621714756e-05, "loss": 0.0483, "step": 8050 }, { "grad_norm": 0.34163832664489746, "learning_rate": 6.96371325503941e-05, "loss": 0.0381, "step": 8060 }, { "grad_norm": 0.3858708143234253, "learning_rate": 6.956107519647014e-05, "loss": 0.0375, "step": 8070 }, { "grad_norm": 0.5838449001312256, "learning_rate": 6.94849643633135e-05, "loss": 0.0674, "step": 8080 }, { "grad_norm": 0.3087199330329895, "learning_rate": 6.940880025900834e-05, "loss": 0.0372, "step": 8090 }, { "grad_norm": 0.29672276973724365, "learning_rate": 6.933258309178438e-05, "loss": 0.0393, "step": 8100 }, { "grad_norm": 0.4190327525138855, "learning_rate": 6.925631307001646e-05, "loss": 0.0491, "step": 8110 }, { "grad_norm": 0.43032583594322205, "learning_rate": 6.91799904022239e-05, "loss": 0.0469, "step": 8120 }, { "grad_norm": 0.44272416830062866, "learning_rate": 6.910361529706997e-05, "loss": 0.0493, "step": 8130 }, { "grad_norm": 0.3027600049972534, "learning_rate": 6.902718796336131e-05, "loss": 0.044, "step": 8140 }, { "grad_norm": 0.3819424510002136, "learning_rate": 6.895070861004729e-05, "loss": 0.0365, "step": 8150 }, { "grad_norm": 0.3146205246448517, "learning_rate": 6.887417744621956e-05, "loss": 0.0394, "step": 8160 }, { "grad_norm": 0.29338547587394714, "learning_rate": 6.87975946811114e-05, "loss": 0.0317, "step": 8170 }, { "grad_norm": 0.4764527976512909, "learning_rate": 6.872096052409718e-05, "loss": 0.033, "step": 8180 }, { "grad_norm": 0.40927112102508545, "learning_rate": 6.864427518469174e-05, "loss": 0.0285, "step": 8190 }, { "grad_norm": 0.24311579763889313, "learning_rate": 6.856753887254986e-05, "loss": 0.0378, "step": 8200 }, { "grad_norm": 0.29087042808532715, "learning_rate": 6.849075179746572e-05, "loss": 0.0233, "step": 8210 }, { "grad_norm": 0.26628559827804565, "learning_rate": 6.841391416937221e-05, "loss": 0.0351, "step": 8220 }, { "grad_norm": 0.2750670611858368, "learning_rate": 6.833702619834053e-05, "loss": 0.0337, "step": 8230 }, { "grad_norm": 0.3893817961215973, "learning_rate": 6.82600880945794e-05, "loss": 0.0447, "step": 8240 }, { "grad_norm": 0.41692882776260376, "learning_rate": 6.818310006843468e-05, "loss": 0.055, "step": 8250 }, { "grad_norm": 0.27028346061706543, "learning_rate": 6.810606233038868e-05, "loss": 0.0259, "step": 8260 }, { "grad_norm": 0.2972779870033264, "learning_rate": 6.802897509105966e-05, "loss": 0.04, "step": 8270 }, { "grad_norm": 0.33484312891960144, "learning_rate": 6.79518385612012e-05, "loss": 0.0441, "step": 8280 }, { "grad_norm": 0.5617392659187317, "learning_rate": 6.787465295170157e-05, "loss": 0.0283, "step": 8290 }, { "grad_norm": 0.2714959979057312, "learning_rate": 6.779741847358332e-05, "loss": 0.0478, "step": 8300 }, { "grad_norm": 0.3243560194969177, "learning_rate": 6.772013533800256e-05, "loss": 0.0245, "step": 8310 }, { "grad_norm": 0.45723798871040344, "learning_rate": 6.764280375624843e-05, "loss": 0.0255, "step": 8320 }, { "grad_norm": 0.39463210105895996, "learning_rate": 6.756542393974252e-05, "loss": 0.0414, "step": 8330 }, { "grad_norm": 0.18346337974071503, "learning_rate": 6.748799610003828e-05, "loss": 0.0291, "step": 8340 }, { "grad_norm": 0.44755497574806213, "learning_rate": 6.741052044882048e-05, "loss": 0.032, "step": 8350 }, { "grad_norm": 0.4502626955509186, "learning_rate": 6.73329971979046e-05, "loss": 0.0344, "step": 8360 }, { "grad_norm": 0.44634199142456055, "learning_rate": 6.725542655923625e-05, "loss": 0.0326, "step": 8370 }, { "grad_norm": 0.4080137312412262, "learning_rate": 6.717780874489057e-05, "loss": 0.0301, "step": 8380 }, { "grad_norm": 0.19677186012268066, "learning_rate": 6.710014396707172e-05, "loss": 0.031, "step": 8390 }, { "grad_norm": 0.3140607476234436, "learning_rate": 6.702243243811221e-05, "loss": 0.0344, "step": 8400 }, { "grad_norm": 0.2361263930797577, "learning_rate": 6.694467437047244e-05, "loss": 0.0346, "step": 8410 }, { "grad_norm": 0.3081178665161133, "learning_rate": 6.686686997673997e-05, "loss": 0.0268, "step": 8420 }, { "grad_norm": 0.4046136140823364, "learning_rate": 6.678901946962903e-05, "loss": 0.0369, "step": 8430 }, { "grad_norm": 0.22517940402030945, "learning_rate": 6.671112306197996e-05, "loss": 0.036, "step": 8440 }, { "grad_norm": 0.4145694673061371, "learning_rate": 6.663318096675854e-05, "loss": 0.0323, "step": 8450 }, { "grad_norm": 0.42365220189094543, "learning_rate": 6.655519339705552e-05, "loss": 0.0427, "step": 8460 }, { "grad_norm": 0.30556055903434753, "learning_rate": 6.647716056608588e-05, "loss": 0.0362, "step": 8470 }, { "grad_norm": 0.37867313623428345, "learning_rate": 6.639908268718843e-05, "loss": 0.0262, "step": 8480 }, { "grad_norm": 0.465231716632843, "learning_rate": 6.632095997382514e-05, "loss": 0.0354, "step": 8490 }, { "grad_norm": 0.45221492648124695, "learning_rate": 6.624279263958047e-05, "loss": 0.0395, "step": 8500 }, { "grad_norm": 0.30612292885780334, "learning_rate": 6.616458089816097e-05, "loss": 0.0341, "step": 8510 }, { "grad_norm": 0.30967992544174194, "learning_rate": 6.608632496339454e-05, "loss": 0.0237, "step": 8520 }, { "grad_norm": 0.3658648133277893, "learning_rate": 6.600802504922988e-05, "loss": 0.0476, "step": 8530 }, { "grad_norm": 0.38406503200531006, "learning_rate": 6.592968136973604e-05, "loss": 0.0351, "step": 8540 }, { "grad_norm": 0.2732188403606415, "learning_rate": 6.585129413910159e-05, "loss": 0.0365, "step": 8550 }, { "grad_norm": 0.34742051362991333, "learning_rate": 6.577286357163424e-05, "loss": 0.0314, "step": 8560 }, { "grad_norm": 0.5613582134246826, "learning_rate": 6.569438988176018e-05, "loss": 0.0582, "step": 8570 }, { "grad_norm": 0.39616453647613525, "learning_rate": 6.561587328402347e-05, "loss": 0.0417, "step": 8580 }, { "grad_norm": 0.4446186423301697, "learning_rate": 6.553731399308549e-05, "loss": 0.0389, "step": 8590 }, { "grad_norm": 0.30579307675361633, "learning_rate": 6.545871222372436e-05, "loss": 0.0312, "step": 8600 }, { "grad_norm": 0.6415376663208008, "learning_rate": 6.538006819083426e-05, "loss": 0.0292, "step": 8610 }, { "grad_norm": 0.5312392115592957, "learning_rate": 6.530138210942505e-05, "loss": 0.0342, "step": 8620 }, { "grad_norm": 0.4598277509212494, "learning_rate": 6.522265419462141e-05, "loss": 0.0448, "step": 8630 }, { "grad_norm": 0.4067436158657074, "learning_rate": 6.514388466166248e-05, "loss": 0.0409, "step": 8640 }, { "grad_norm": 0.3071712553501129, "learning_rate": 6.506507372590119e-05, "loss": 0.035, "step": 8650 }, { "grad_norm": 0.33204761147499084, "learning_rate": 6.498622160280355e-05, "loss": 0.0431, "step": 8660 }, { "grad_norm": 0.36536115407943726, "learning_rate": 6.490732850794832e-05, "loss": 0.0308, "step": 8670 }, { "grad_norm": 0.2782830595970154, "learning_rate": 6.482839465702616e-05, "loss": 0.0287, "step": 8680 }, { "grad_norm": 0.36908599734306335, "learning_rate": 6.474942026583923e-05, "loss": 0.0318, "step": 8690 }, { "grad_norm": 0.4494503438472748, "learning_rate": 6.467040555030052e-05, "loss": 0.0305, "step": 8700 }, { "grad_norm": 0.30583712458610535, "learning_rate": 6.459135072643321e-05, "loss": 0.0466, "step": 8710 }, { "grad_norm": 0.7435348033905029, "learning_rate": 6.451225601037019e-05, "loss": 0.0488, "step": 8720 }, { "grad_norm": 0.25439444184303284, "learning_rate": 6.443312161835338e-05, "loss": 0.0305, "step": 8730 }, { "grad_norm": 0.2972891628742218, "learning_rate": 6.43539477667332e-05, "loss": 0.0326, "step": 8740 }, { "grad_norm": 0.3128562271595001, "learning_rate": 6.427473467196793e-05, "loss": 0.0388, "step": 8750 }, { "grad_norm": 0.34789153933525085, "learning_rate": 6.419548255062315e-05, "loss": 0.0312, "step": 8760 }, { "grad_norm": 0.27936166524887085, "learning_rate": 6.411619161937112e-05, "loss": 0.0352, "step": 8770 }, { "grad_norm": 0.39509639143943787, "learning_rate": 6.403686209499022e-05, "loss": 0.0436, "step": 8780 }, { "grad_norm": 0.2747631072998047, "learning_rate": 6.395749419436437e-05, "loss": 0.0256, "step": 8790 }, { "grad_norm": 0.46139782667160034, "learning_rate": 6.387808813448234e-05, "loss": 0.0346, "step": 8800 }, { "grad_norm": 0.14183153212070465, "learning_rate": 6.37986441324373e-05, "loss": 0.0358, "step": 8810 }, { "grad_norm": 0.35424017906188965, "learning_rate": 6.37191624054261e-05, "loss": 0.0406, "step": 8820 }, { "grad_norm": 0.40496084094047546, "learning_rate": 6.363964317074872e-05, "loss": 0.0342, "step": 8830 }, { "grad_norm": 0.41786879301071167, "learning_rate": 6.356008664580776e-05, "loss": 0.034, "step": 8840 }, { "grad_norm": 0.35758182406425476, "learning_rate": 6.348049304810771e-05, "loss": 0.0292, "step": 8850 }, { "grad_norm": 0.343021035194397, "learning_rate": 6.340086259525442e-05, "loss": 0.0257, "step": 8860 }, { "grad_norm": 0.5082806944847107, "learning_rate": 6.332119550495448e-05, "loss": 0.0299, "step": 8870 }, { "grad_norm": 0.3564285933971405, "learning_rate": 6.324149199501473e-05, "loss": 0.0319, "step": 8880 }, { "grad_norm": 0.3154931664466858, "learning_rate": 6.316175228334146e-05, "loss": 0.0445, "step": 8890 }, { "grad_norm": 0.45812296867370605, "learning_rate": 6.308197658794003e-05, "loss": 0.0316, "step": 8900 }, { "grad_norm": 0.17545326054096222, "learning_rate": 6.300216512691417e-05, "loss": 0.0265, "step": 8910 }, { "grad_norm": 0.47893226146698, "learning_rate": 6.292231811846532e-05, "loss": 0.0332, "step": 8920 }, { "grad_norm": 0.2554419934749603, "learning_rate": 6.284243578089217e-05, "loss": 0.0255, "step": 8930 }, { "grad_norm": 0.21513232588768005, "learning_rate": 6.276251833258999e-05, "loss": 0.0316, "step": 8940 }, { "grad_norm": 0.20232705771923065, "learning_rate": 6.268256599205003e-05, "loss": 0.0327, "step": 8950 }, { "grad_norm": 0.43983879685401917, "learning_rate": 6.260257897785892e-05, "loss": 0.0258, "step": 8960 }, { "grad_norm": 0.281688928604126, "learning_rate": 6.252255750869811e-05, "loss": 0.0257, "step": 8970 }, { "grad_norm": 0.1872902810573578, "learning_rate": 6.244250180334325e-05, "loss": 0.0279, "step": 8980 }, { "grad_norm": 0.5532070994377136, "learning_rate": 6.236241208066356e-05, "loss": 0.0536, "step": 8990 }, { "grad_norm": 0.22434480488300323, "learning_rate": 6.228228855962133e-05, "loss": 0.0311, "step": 9000 }, { "grad_norm": 0.24825334548950195, "learning_rate": 6.220213145927115e-05, "loss": 0.027, "step": 9010 }, { "grad_norm": 0.4586460590362549, "learning_rate": 6.212194099875951e-05, "loss": 0.037, "step": 9020 }, { "grad_norm": 0.2770425081253052, "learning_rate": 6.204171739732405e-05, "loss": 0.0362, "step": 9030 }, { "grad_norm": 0.30298352241516113, "learning_rate": 6.196146087429303e-05, "loss": 0.0338, "step": 9040 }, { "grad_norm": 0.3805689811706543, "learning_rate": 6.188117164908474e-05, "loss": 0.0383, "step": 9050 }, { "grad_norm": 0.49097540974617004, "learning_rate": 6.180084994120684e-05, "loss": 0.0498, "step": 9060 }, { "grad_norm": 0.23234422504901886, "learning_rate": 6.17204959702558e-05, "loss": 0.0257, "step": 9070 }, { "grad_norm": 0.5593923926353455, "learning_rate": 6.164010995591635e-05, "loss": 0.0252, "step": 9080 }, { "grad_norm": 0.2584875822067261, "learning_rate": 6.155969211796076e-05, "loss": 0.0428, "step": 9090 }, { "grad_norm": 0.47834184765815735, "learning_rate": 6.147924267624829e-05, "loss": 0.0372, "step": 9100 }, { "grad_norm": 0.5277264714241028, "learning_rate": 6.13987618507247e-05, "loss": 0.0414, "step": 9110 }, { "grad_norm": 0.44414371252059937, "learning_rate": 6.131824986142147e-05, "loss": 0.0392, "step": 9120 }, { "grad_norm": 0.37389329075813293, "learning_rate": 6.123770692845529e-05, "loss": 0.0348, "step": 9130 }, { "grad_norm": 0.39935240149497986, "learning_rate": 6.11571332720275e-05, "loss": 0.0314, "step": 9140 }, { "grad_norm": 0.368661105632782, "learning_rate": 6.107652911242336e-05, "loss": 0.0385, "step": 9150 }, { "grad_norm": 0.2869284749031067, "learning_rate": 6.0995894670011586e-05, "loss": 0.0284, "step": 9160 }, { "grad_norm": 0.36483851075172424, "learning_rate": 6.091523016524368e-05, "loss": 0.0383, "step": 9170 }, { "grad_norm": 0.2807167172431946, "learning_rate": 6.083453581865328e-05, "loss": 0.0295, "step": 9180 }, { "grad_norm": 0.17450042068958282, "learning_rate": 6.075381185085568e-05, "loss": 0.0277, "step": 9190 }, { "grad_norm": 0.2801981568336487, "learning_rate": 6.067305848254709e-05, "loss": 0.0381, "step": 9200 }, { "grad_norm": 0.3023240864276886, "learning_rate": 6.059227593450418e-05, "loss": 0.042, "step": 9210 }, { "grad_norm": 0.18939031660556793, "learning_rate": 6.051146442758333e-05, "loss": 0.042, "step": 9220 }, { "grad_norm": 0.3736701011657715, "learning_rate": 6.043062418272012e-05, "loss": 0.0306, "step": 9230 }, { "grad_norm": 0.22779077291488647, "learning_rate": 6.0349755420928666e-05, "loss": 0.0374, "step": 9240 }, { "grad_norm": 0.6783238053321838, "learning_rate": 6.0268858363301105e-05, "loss": 0.0311, "step": 9250 }, { "grad_norm": 0.3538046181201935, "learning_rate": 6.018793323100689e-05, "loss": 0.0276, "step": 9260 }, { "grad_norm": 0.34534555673599243, "learning_rate": 6.0106980245292255e-05, "loss": 0.0325, "step": 9270 }, { "grad_norm": 0.3202175498008728, "learning_rate": 6.002599962747957e-05, "loss": 0.0283, "step": 9280 }, { "grad_norm": 0.4542574882507324, "learning_rate": 5.994499159896673e-05, "loss": 0.0485, "step": 9290 }, { "grad_norm": 0.20237502455711365, "learning_rate": 5.9863956381226607e-05, "loss": 0.0369, "step": 9300 }, { "grad_norm": 0.6103059649467468, "learning_rate": 5.9782894195806394e-05, "loss": 0.0349, "step": 9310 }, { "grad_norm": 0.5782369375228882, "learning_rate": 5.9701805264327004e-05, "loss": 0.049, "step": 9320 }, { "grad_norm": 0.5471524000167847, "learning_rate": 5.96206898084825e-05, "loss": 0.0291, "step": 9330 }, { "grad_norm": 0.24528686702251434, "learning_rate": 5.953954805003942e-05, "loss": 0.0345, "step": 9340 }, { "grad_norm": 0.36265459656715393, "learning_rate": 5.945838021083623e-05, "loss": 0.0276, "step": 9350 }, { "grad_norm": 0.2779427170753479, "learning_rate": 5.9377186512782714e-05, "loss": 0.0258, "step": 9360 }, { "grad_norm": 0.22113096714019775, "learning_rate": 5.929596717785935e-05, "loss": 0.0452, "step": 9370 }, { "grad_norm": 0.31225425004959106, "learning_rate": 5.921472242811668e-05, "loss": 0.0367, "step": 9380 }, { "grad_norm": 0.2538183033466339, "learning_rate": 5.913345248567475e-05, "loss": 0.0314, "step": 9390 }, { "grad_norm": 0.2832610309123993, "learning_rate": 5.905215757272248e-05, "loss": 0.0377, "step": 9400 }, { "grad_norm": 0.22567032277584076, "learning_rate": 5.897083791151706e-05, "loss": 0.0263, "step": 9410 }, { "grad_norm": 0.394778847694397, "learning_rate": 5.888949372438336e-05, "loss": 0.0348, "step": 9420 }, { "grad_norm": 0.30844902992248535, "learning_rate": 5.8808125233713255e-05, "loss": 0.0237, "step": 9430 }, { "grad_norm": 0.3853760361671448, "learning_rate": 5.872673266196509e-05, "loss": 0.0273, "step": 9440 }, { "grad_norm": 0.2399638146162033, "learning_rate": 5.864531623166305e-05, "loss": 0.0259, "step": 9450 }, { "grad_norm": 0.18106147646903992, "learning_rate": 5.856387616539656e-05, "loss": 0.0208, "step": 9460 }, { "grad_norm": 0.36653029918670654, "learning_rate": 5.848241268581967e-05, "loss": 0.0525, "step": 9470 }, { "grad_norm": 0.2835184335708618, "learning_rate": 5.840092601565037e-05, "loss": 0.0266, "step": 9480 }, { "grad_norm": 1.3431917428970337, "learning_rate": 5.8319416377670144e-05, "loss": 0.0397, "step": 9490 }, { "grad_norm": 0.3711817264556885, "learning_rate": 5.82378839947232e-05, "loss": 0.0369, "step": 9500 }, { "grad_norm": 0.20385318994522095, "learning_rate": 5.815632908971599e-05, "loss": 0.0338, "step": 9510 }, { "grad_norm": 0.2646554708480835, "learning_rate": 5.80747518856165e-05, "loss": 0.032, "step": 9520 }, { "grad_norm": 0.182133287191391, "learning_rate": 5.799315260545367e-05, "loss": 0.032, "step": 9530 }, { "grad_norm": 0.5178720355033875, "learning_rate": 5.791153147231686e-05, "loss": 0.0311, "step": 9540 }, { "grad_norm": 0.23337844014167786, "learning_rate": 5.782988870935509e-05, "loss": 0.0255, "step": 9550 }, { "grad_norm": 0.3603869080543518, "learning_rate": 5.774822453977657e-05, "loss": 0.0344, "step": 9560 }, { "grad_norm": 1.0083812475204468, "learning_rate": 5.7666539186848036e-05, "loss": 0.0573, "step": 9570 }, { "grad_norm": 0.477022260427475, "learning_rate": 5.758483287389411e-05, "loss": 0.0278, "step": 9580 }, { "grad_norm": 0.5770471692085266, "learning_rate": 5.7503105824296735e-05, "loss": 0.0404, "step": 9590 }, { "grad_norm": 0.2824535071849823, "learning_rate": 5.742135826149453e-05, "loss": 0.034, "step": 9600 }, { "grad_norm": 0.2005973607301712, "learning_rate": 5.7339590408982223e-05, "loss": 0.0244, "step": 9610 }, { "grad_norm": 0.31074148416519165, "learning_rate": 5.725780249031e-05, "loss": 0.0326, "step": 9620 }, { "grad_norm": 0.2513398230075836, "learning_rate": 5.717599472908292e-05, "loss": 0.0375, "step": 9630 }, { "grad_norm": 0.3207153081893921, "learning_rate": 5.7094167348960237e-05, "loss": 0.0282, "step": 9640 }, { "grad_norm": 0.18063382804393768, "learning_rate": 5.7012320573654945e-05, "loss": 0.0255, "step": 9650 }, { "grad_norm": 0.24532638490200043, "learning_rate": 5.693045462693295e-05, "loss": 0.0325, "step": 9660 }, { "grad_norm": 0.3828868865966797, "learning_rate": 5.684856973261266e-05, "loss": 0.0241, "step": 9670 }, { "grad_norm": 0.21055418252944946, "learning_rate": 5.6766666114564215e-05, "loss": 0.0202, "step": 9680 }, { "grad_norm": 0.19171160459518433, "learning_rate": 5.668474399670899e-05, "loss": 0.0258, "step": 9690 }, { "grad_norm": 0.3057857155799866, "learning_rate": 5.660280360301896e-05, "loss": 0.024, "step": 9700 }, { "grad_norm": 0.24466152489185333, "learning_rate": 5.652084515751599e-05, "loss": 0.0278, "step": 9710 }, { "grad_norm": 0.40165966749191284, "learning_rate": 5.643886888427137e-05, "loss": 0.0287, "step": 9720 }, { "grad_norm": 0.6342644095420837, "learning_rate": 5.6356875007405074e-05, "loss": 0.0358, "step": 9730 }, { "grad_norm": 0.31248196959495544, "learning_rate": 5.627486375108525e-05, "loss": 0.0267, "step": 9740 }, { "grad_norm": 0.352142333984375, "learning_rate": 5.619283533952754e-05, "loss": 0.0312, "step": 9750 }, { "grad_norm": 0.2738164961338043, "learning_rate": 5.6110789996994474e-05, "loss": 0.0324, "step": 9760 }, { "grad_norm": 0.40617287158966064, "learning_rate": 5.602872794779491e-05, "loss": 0.0319, "step": 9770 }, { "grad_norm": 0.5647812485694885, "learning_rate": 5.594664941628334e-05, "loss": 0.0308, "step": 9780 }, { "grad_norm": 0.22140783071517944, "learning_rate": 5.5864554626859324e-05, "loss": 0.0243, "step": 9790 }, { "grad_norm": 0.4303944706916809, "learning_rate": 5.578244380396691e-05, "loss": 0.0289, "step": 9800 }, { "grad_norm": 0.5359013080596924, "learning_rate": 5.570031717209394e-05, "loss": 0.0347, "step": 9810 }, { "grad_norm": 0.45951947569847107, "learning_rate": 5.561817495577147e-05, "loss": 0.0373, "step": 9820 }, { "grad_norm": 0.2878282070159912, "learning_rate": 5.5536017379573215e-05, "loss": 0.0253, "step": 9830 }, { "grad_norm": 0.28325560688972473, "learning_rate": 5.545384466811483e-05, "loss": 0.0466, "step": 9840 }, { "grad_norm": 0.30453234910964966, "learning_rate": 5.5371657046053384e-05, "loss": 0.042, "step": 9850 }, { "grad_norm": 0.3731170892715454, "learning_rate": 5.528945473808669e-05, "loss": 0.0286, "step": 9860 }, { "grad_norm": 0.3516085147857666, "learning_rate": 5.520723796895272e-05, "loss": 0.0248, "step": 9870 }, { "grad_norm": 0.4815094470977783, "learning_rate": 5.512500696342897e-05, "loss": 0.0362, "step": 9880 }, { "grad_norm": 0.1732551008462906, "learning_rate": 5.504276194633188e-05, "loss": 0.0259, "step": 9890 }, { "grad_norm": 0.7045866250991821, "learning_rate": 5.49605031425162e-05, "loss": 0.0338, "step": 9900 }, { "grad_norm": 0.18000729382038116, "learning_rate": 5.487823077687434e-05, "loss": 0.031, "step": 9910 }, { "grad_norm": 0.31892889738082886, "learning_rate": 5.4795945074335806e-05, "loss": 0.0348, "step": 9920 }, { "grad_norm": 0.3102439343929291, "learning_rate": 5.471364625986657e-05, "loss": 0.0254, "step": 9930 }, { "grad_norm": 0.2934859097003937, "learning_rate": 5.463133455846845e-05, "loss": 0.0288, "step": 9940 }, { "grad_norm": 0.2539833188056946, "learning_rate": 5.4549010195178505e-05, "loss": 0.0206, "step": 9950 }, { "grad_norm": 0.2775578498840332, "learning_rate": 5.446667339506838e-05, "loss": 0.0269, "step": 9960 }, { "grad_norm": 0.2690306007862091, "learning_rate": 5.4384324383243756e-05, "loss": 0.0382, "step": 9970 }, { "grad_norm": 0.29434746503829956, "learning_rate": 5.430196338484368e-05, "loss": 0.0241, "step": 9980 }, { "grad_norm": 0.3495558202266693, "learning_rate": 5.4219590625039975e-05, "loss": 0.0295, "step": 9990 }, { "grad_norm": 0.529267430305481, "learning_rate": 5.413720632903664e-05, "loss": 0.0332, "step": 10000 }, { "grad_norm": 0.3043668568134308, "learning_rate": 5.405481072206917e-05, "loss": 0.0344, "step": 10010 }, { "grad_norm": 0.2161196917295456, "learning_rate": 5.397240402940402e-05, "loss": 0.034, "step": 10020 }, { "grad_norm": 0.2535768747329712, "learning_rate": 5.388998647633794e-05, "loss": 0.024, "step": 10030 }, { "grad_norm": 0.43511658906936646, "learning_rate": 5.380755828819737e-05, "loss": 0.0362, "step": 10040 }, { "grad_norm": 0.29506123065948486, "learning_rate": 5.3725119690337846e-05, "loss": 0.0328, "step": 10050 }, { "grad_norm": 0.7222499847412109, "learning_rate": 5.3642670908143324e-05, "loss": 0.0241, "step": 10060 }, { "grad_norm": 0.17279590666294098, "learning_rate": 5.356021216702562e-05, "loss": 0.0166, "step": 10070 }, { "grad_norm": 0.16380324959754944, "learning_rate": 5.347774369242381e-05, "loss": 0.0254, "step": 10080 }, { "grad_norm": 0.5014934539794922, "learning_rate": 5.3395265709803545e-05, "loss": 0.0271, "step": 10090 }, { "grad_norm": 0.2136426419019699, "learning_rate": 5.331277844465647e-05, "loss": 0.0224, "step": 10100 }, { "grad_norm": 0.14739011228084564, "learning_rate": 5.323028212249963e-05, "loss": 0.0172, "step": 10110 }, { "grad_norm": 0.19700703024864197, "learning_rate": 5.314777696887481e-05, "loss": 0.0243, "step": 10120 }, { "grad_norm": 0.46293023228645325, "learning_rate": 5.306526320934796e-05, "loss": 0.0381, "step": 10130 }, { "grad_norm": 0.3641117811203003, "learning_rate": 5.298274106950854e-05, "loss": 0.032, "step": 10140 }, { "grad_norm": 0.3949614465236664, "learning_rate": 5.290021077496893e-05, "loss": 0.034, "step": 10150 }, { "grad_norm": 0.30592426657676697, "learning_rate": 5.2817672551363816e-05, "loss": 0.0295, "step": 10160 }, { "grad_norm": 0.2354503571987152, "learning_rate": 5.273512662434952e-05, "loss": 0.0353, "step": 10170 }, { "grad_norm": 0.2648562788963318, "learning_rate": 5.265257321960349e-05, "loss": 0.0412, "step": 10180 }, { "grad_norm": 0.48530954122543335, "learning_rate": 5.257001256282357e-05, "loss": 0.0441, "step": 10190 }, { "grad_norm": 0.3964565396308899, "learning_rate": 5.248744487972742e-05, "loss": 0.0339, "step": 10200 }, { "grad_norm": 0.2062573879957199, "learning_rate": 5.240487039605196e-05, "loss": 0.0354, "step": 10210 }, { "grad_norm": 0.3242119252681732, "learning_rate": 5.232228933755267e-05, "loss": 0.0228, "step": 10220 }, { "grad_norm": 0.33489319682121277, "learning_rate": 5.2239701930003006e-05, "loss": 0.0301, "step": 10230 }, { "grad_norm": 0.2048545479774475, "learning_rate": 5.215710839919379e-05, "loss": 0.0336, "step": 10240 }, { "grad_norm": 0.26481661200523376, "learning_rate": 5.207450897093257e-05, "loss": 0.0339, "step": 10250 }, { "grad_norm": 0.16015321016311646, "learning_rate": 5.1991903871043046e-05, "loss": 0.0258, "step": 10260 }, { "grad_norm": 0.2631183862686157, "learning_rate": 5.190929332536439e-05, "loss": 0.0262, "step": 10270 }, { "grad_norm": 0.2275284081697464, "learning_rate": 5.182667755975071e-05, "loss": 0.0342, "step": 10280 }, { "grad_norm": 0.27065250277519226, "learning_rate": 5.1744056800070315e-05, "loss": 0.0262, "step": 10290 }, { "grad_norm": 0.40078839659690857, "learning_rate": 5.166143127220524e-05, "loss": 0.0266, "step": 10300 }, { "grad_norm": 0.513117253780365, "learning_rate": 5.1578801202050485e-05, "loss": 0.0209, "step": 10310 }, { "grad_norm": 0.33915454149246216, "learning_rate": 5.149616681551355e-05, "loss": 0.0253, "step": 10320 }, { "grad_norm": 0.2879283130168915, "learning_rate": 5.141352833851367e-05, "loss": 0.0295, "step": 10330 }, { "grad_norm": 0.26950517296791077, "learning_rate": 5.1330885996981285e-05, "loss": 0.0259, "step": 10340 }, { "grad_norm": 0.3174979090690613, "learning_rate": 5.124824001685741e-05, "loss": 0.0314, "step": 10350 }, { "grad_norm": 0.13502341508865356, "learning_rate": 5.116559062409298e-05, "loss": 0.0281, "step": 10360 }, { "grad_norm": 0.2029581516981125, "learning_rate": 5.10829380446483e-05, "loss": 0.0265, "step": 10370 }, { "grad_norm": 0.4445574879646301, "learning_rate": 5.100028250449235e-05, "loss": 0.0426, "step": 10380 }, { "grad_norm": 0.2911873459815979, "learning_rate": 5.0917624229602234e-05, "loss": 0.0293, "step": 10390 }, { "grad_norm": 0.2548791170120239, "learning_rate": 5.0834963445962524e-05, "loss": 0.0301, "step": 10400 }, { "grad_norm": 0.2480202168226242, "learning_rate": 5.075230037956461e-05, "loss": 0.0255, "step": 10410 }, { "grad_norm": 0.2780553698539734, "learning_rate": 5.0669635256406213e-05, "loss": 0.0354, "step": 10420 }, { "grad_norm": 0.3311081826686859, "learning_rate": 5.058696830249058e-05, "loss": 0.0285, "step": 10430 }, { "grad_norm": 0.3468809127807617, "learning_rate": 5.050429974382602e-05, "loss": 0.0199, "step": 10440 }, { "grad_norm": 0.4322201609611511, "learning_rate": 5.042162980642523e-05, "loss": 0.0299, "step": 10450 }, { "grad_norm": 0.3680262863636017, "learning_rate": 5.033895871630462e-05, "loss": 0.0323, "step": 10460 }, { "grad_norm": 0.3636588454246521, "learning_rate": 5.025628669948386e-05, "loss": 0.0326, "step": 10470 }, { "grad_norm": 0.32991665601730347, "learning_rate": 5.017361398198502e-05, "loss": 0.0257, "step": 10480 }, { "grad_norm": 0.3385353088378906, "learning_rate": 5.009094078983221e-05, "loss": 0.0313, "step": 10490 }, { "grad_norm": 0.21950207650661469, "learning_rate": 5.000826734905073e-05, "loss": 0.0192, "step": 10500 }, { "grad_norm": 0.5016471147537231, "learning_rate": 4.9925593885666645e-05, "loss": 0.0396, "step": 10510 }, { "grad_norm": 0.36445486545562744, "learning_rate": 4.984292062570602e-05, "loss": 0.0204, "step": 10520 }, { "grad_norm": 0.20073537528514862, "learning_rate": 4.976024779519442e-05, "loss": 0.037, "step": 10530 }, { "grad_norm": 0.36668631434440613, "learning_rate": 4.9677575620156194e-05, "loss": 0.0222, "step": 10540 }, { "grad_norm": 0.3067803382873535, "learning_rate": 4.959490432661391e-05, "loss": 0.0241, "step": 10550 }, { "grad_norm": 0.41551700234413147, "learning_rate": 4.9512234140587726e-05, "loss": 0.0202, "step": 10560 }, { "grad_norm": 0.1831006109714508, "learning_rate": 4.942956528809477e-05, "loss": 0.0268, "step": 10570 }, { "grad_norm": 0.2666257917881012, "learning_rate": 4.934689799514854e-05, "loss": 0.0361, "step": 10580 }, { "grad_norm": 0.247621551156044, "learning_rate": 4.926423248775827e-05, "loss": 0.0311, "step": 10590 }, { "grad_norm": 0.23860666155815125, "learning_rate": 4.918156899192826e-05, "loss": 0.0406, "step": 10600 }, { "grad_norm": 0.6485392451286316, "learning_rate": 4.909890773365738e-05, "loss": 0.0434, "step": 10610 }, { "grad_norm": 0.4776696562767029, "learning_rate": 4.9016248938938344e-05, "loss": 0.0393, "step": 10620 }, { "grad_norm": 0.206770658493042, "learning_rate": 4.8933592833757156e-05, "loss": 0.029, "step": 10630 }, { "grad_norm": 0.3446054458618164, "learning_rate": 4.8850939644092435e-05, "loss": 0.04, "step": 10640 }, { "grad_norm": 0.19238072633743286, "learning_rate": 4.876828959591485e-05, "loss": 0.0256, "step": 10650 }, { "grad_norm": 0.2756827771663666, "learning_rate": 4.8685642915186474e-05, "loss": 0.0321, "step": 10660 }, { "grad_norm": 0.5545204877853394, "learning_rate": 4.860299982786018e-05, "loss": 0.035, "step": 10670 }, { "grad_norm": 0.33314889669418335, "learning_rate": 4.852036055987901e-05, "loss": 0.0243, "step": 10680 }, { "grad_norm": 0.21042852103710175, "learning_rate": 4.843772533717558e-05, "loss": 0.0232, "step": 10690 }, { "grad_norm": 0.4332270920276642, "learning_rate": 4.835509438567142e-05, "loss": 0.046, "step": 10700 }, { "grad_norm": 0.2374183088541031, "learning_rate": 4.827246793127639e-05, "loss": 0.0253, "step": 10710 }, { "grad_norm": 0.3086862862110138, "learning_rate": 4.818984619988807e-05, "loss": 0.0355, "step": 10720 }, { "grad_norm": 0.34332355856895447, "learning_rate": 4.810722941739115e-05, "loss": 0.0301, "step": 10730 }, { "grad_norm": 0.45614132285118103, "learning_rate": 4.8024617809656684e-05, "loss": 0.0271, "step": 10740 }, { "grad_norm": 0.33681970834732056, "learning_rate": 4.794201160254171e-05, "loss": 0.0274, "step": 10750 }, { "grad_norm": 0.3248555064201355, "learning_rate": 4.785941102188844e-05, "loss": 0.0399, "step": 10760 }, { "grad_norm": 0.5792879462242126, "learning_rate": 4.7776816293523686e-05, "loss": 0.037, "step": 10770 }, { "grad_norm": 0.5702614188194275, "learning_rate": 4.769422764325832e-05, "loss": 0.0308, "step": 10780 }, { "grad_norm": 0.34837472438812256, "learning_rate": 4.76116452968865e-05, "loss": 0.0331, "step": 10790 }, { "grad_norm": 0.34685492515563965, "learning_rate": 4.752906948018525e-05, "loss": 0.0308, "step": 10800 }, { "grad_norm": 0.514130711555481, "learning_rate": 4.7446500418913684e-05, "loss": 0.0281, "step": 10810 }, { "grad_norm": 0.23345115780830383, "learning_rate": 4.736393833881247e-05, "loss": 0.0386, "step": 10820 }, { "grad_norm": 0.20355211198329926, "learning_rate": 4.7281383465603194e-05, "loss": 0.0385, "step": 10830 }, { "grad_norm": 0.15992626547813416, "learning_rate": 4.71988360249877e-05, "loss": 0.0332, "step": 10840 }, { "grad_norm": 0.5549060702323914, "learning_rate": 4.7116296242647554e-05, "loss": 0.0375, "step": 10850 }, { "grad_norm": 0.5806261897087097, "learning_rate": 4.703376434424336e-05, "loss": 0.0316, "step": 10860 }, { "grad_norm": 0.2695363461971283, "learning_rate": 4.695124055541421e-05, "loss": 0.0321, "step": 10870 }, { "grad_norm": 0.2607054114341736, "learning_rate": 4.6868725101776934e-05, "loss": 0.0382, "step": 10880 }, { "grad_norm": 0.4391457438468933, "learning_rate": 4.678621820892567e-05, "loss": 0.0385, "step": 10890 }, { "grad_norm": 0.3854418694972992, "learning_rate": 4.670372010243111e-05, "loss": 0.0408, "step": 10900 }, { "grad_norm": 0.2532274127006531, "learning_rate": 4.662123100783992e-05, "loss": 0.0291, "step": 10910 }, { "grad_norm": 0.33813488483428955, "learning_rate": 4.653875115067415e-05, "loss": 0.0348, "step": 10920 }, { "grad_norm": 0.226698100566864, "learning_rate": 4.6456280756430545e-05, "loss": 0.0362, "step": 10930 }, { "grad_norm": 0.3152325749397278, "learning_rate": 4.637382005058004e-05, "loss": 0.0267, "step": 10940 }, { "grad_norm": 0.329550176858902, "learning_rate": 4.629136925856705e-05, "loss": 0.0338, "step": 10950 }, { "grad_norm": 0.19500219821929932, "learning_rate": 4.6208928605808895e-05, "loss": 0.026, "step": 10960 }, { "grad_norm": 0.3183884620666504, "learning_rate": 4.612649831769519e-05, "loss": 0.0275, "step": 10970 }, { "grad_norm": 0.1490533947944641, "learning_rate": 4.604407861958715e-05, "loss": 0.0267, "step": 10980 }, { "grad_norm": 0.19505657255649567, "learning_rate": 4.5961669736817114e-05, "loss": 0.0166, "step": 10990 }, { "grad_norm": 0.3448111116886139, "learning_rate": 4.5879271894687814e-05, "loss": 0.0269, "step": 11000 }, { "grad_norm": 0.32773834466934204, "learning_rate": 4.5796885318471826e-05, "loss": 0.0265, "step": 11010 }, { "grad_norm": 0.2204039841890335, "learning_rate": 4.571451023341086e-05, "loss": 0.0274, "step": 11020 }, { "grad_norm": 0.4743645191192627, "learning_rate": 4.563214686471527e-05, "loss": 0.0335, "step": 11030 }, { "grad_norm": 0.43818795680999756, "learning_rate": 4.5549795437563365e-05, "loss": 0.0301, "step": 11040 }, { "grad_norm": 0.24953724443912506, "learning_rate": 4.546745617710081e-05, "loss": 0.0271, "step": 11050 }, { "grad_norm": 0.4423724114894867, "learning_rate": 4.5385129308440014e-05, "loss": 0.0382, "step": 11060 }, { "grad_norm": 0.3067225217819214, "learning_rate": 4.530281505665944e-05, "loss": 0.0245, "step": 11070 }, { "grad_norm": 0.25571057200431824, "learning_rate": 4.5220513646803134e-05, "loss": 0.0307, "step": 11080 }, { "grad_norm": 0.24015256762504578, "learning_rate": 4.513822530388003e-05, "loss": 0.0211, "step": 11090 }, { "grad_norm": 0.20714981853961945, "learning_rate": 4.5055950252863296e-05, "loss": 0.0291, "step": 11100 }, { "grad_norm": 0.2735294997692108, "learning_rate": 4.4973688718689803e-05, "loss": 0.0223, "step": 11110 }, { "grad_norm": 0.26036912202835083, "learning_rate": 4.4891440926259406e-05, "loss": 0.0311, "step": 11120 }, { "grad_norm": 0.2855947017669678, "learning_rate": 4.480920710043443e-05, "loss": 0.0406, "step": 11130 }, { "grad_norm": 0.2160746455192566, "learning_rate": 4.4726987466039044e-05, "loss": 0.0233, "step": 11140 }, { "grad_norm": 0.27769628167152405, "learning_rate": 4.46447822478586e-05, "loss": 0.0246, "step": 11150 }, { "grad_norm": 0.5913590788841248, "learning_rate": 4.4562591670638974e-05, "loss": 0.0343, "step": 11160 }, { "grad_norm": 0.1843281388282776, "learning_rate": 4.4480415959086105e-05, "loss": 0.0232, "step": 11170 }, { "grad_norm": 0.22262164950370789, "learning_rate": 4.439825533786522e-05, "loss": 0.0317, "step": 11180 }, { "grad_norm": 0.38203108310699463, "learning_rate": 4.431611003160035e-05, "loss": 0.0306, "step": 11190 }, { "grad_norm": 0.3752078115940094, "learning_rate": 4.4233980264873636e-05, "loss": 0.0283, "step": 11200 }, { "grad_norm": 0.37995484471321106, "learning_rate": 4.4151866262224684e-05, "loss": 0.0333, "step": 11210 }, { "grad_norm": 0.2199382334947586, "learning_rate": 4.406976824815006e-05, "loss": 0.0279, "step": 11220 }, { "grad_norm": 0.26995688676834106, "learning_rate": 4.3987686447102595e-05, "loss": 0.0338, "step": 11230 }, { "grad_norm": 0.23235087096691132, "learning_rate": 4.3905621083490804e-05, "loss": 0.032, "step": 11240 }, { "grad_norm": 0.30057433247566223, "learning_rate": 4.3823572381678286e-05, "loss": 0.0252, "step": 11250 }, { "grad_norm": 0.27750131487846375, "learning_rate": 4.374154056598301e-05, "loss": 0.0199, "step": 11260 }, { "grad_norm": 0.2086607664823532, "learning_rate": 4.3659525860676845e-05, "loss": 0.0216, "step": 11270 }, { "grad_norm": 0.4005979895591736, "learning_rate": 4.3577528489984854e-05, "loss": 0.0263, "step": 11280 }, { "grad_norm": 0.27407410740852356, "learning_rate": 4.349554867808476e-05, "loss": 0.0231, "step": 11290 }, { "grad_norm": 0.27793145179748535, "learning_rate": 4.34135866491062e-05, "loss": 0.0231, "step": 11300 }, { "grad_norm": 0.5587870478630066, "learning_rate": 4.333164262713022e-05, "loss": 0.0276, "step": 11310 }, { "grad_norm": 0.3061603605747223, "learning_rate": 4.324971683618868e-05, "loss": 0.0311, "step": 11320 }, { "grad_norm": 0.14884373545646667, "learning_rate": 4.316780950026354e-05, "loss": 0.0233, "step": 11330 }, { "grad_norm": 0.4809848368167877, "learning_rate": 4.308592084328637e-05, "loss": 0.029, "step": 11340 }, { "grad_norm": 0.24778184294700623, "learning_rate": 4.3004051089137576e-05, "loss": 0.0382, "step": 11350 }, { "grad_norm": 0.3415316045284271, "learning_rate": 4.292220046164597e-05, "loss": 0.026, "step": 11360 }, { "grad_norm": 0.22430600225925446, "learning_rate": 4.2840369184588035e-05, "loss": 0.0155, "step": 11370 }, { "grad_norm": 0.23974350094795227, "learning_rate": 4.2758557481687345e-05, "loss": 0.0248, "step": 11380 }, { "grad_norm": 0.44596531987190247, "learning_rate": 4.267676557661403e-05, "loss": 0.0322, "step": 11390 }, { "grad_norm": 0.17690689861774445, "learning_rate": 4.2594993692983955e-05, "loss": 0.0341, "step": 11400 }, { "grad_norm": 0.3454243242740631, "learning_rate": 4.251324205435837e-05, "loss": 0.0209, "step": 11410 }, { "grad_norm": 0.2438250035047531, "learning_rate": 4.243151088424312e-05, "loss": 0.0267, "step": 11420 }, { "grad_norm": 0.33163151144981384, "learning_rate": 4.234980040608813e-05, "loss": 0.0252, "step": 11430 }, { "grad_norm": 0.2071181684732437, "learning_rate": 4.22681108432867e-05, "loss": 0.0257, "step": 11440 }, { "grad_norm": 0.3169047236442566, "learning_rate": 4.2186442419174984e-05, "loss": 0.0208, "step": 11450 }, { "grad_norm": 0.35895857214927673, "learning_rate": 4.210479535703133e-05, "loss": 0.0371, "step": 11460 }, { "grad_norm": 0.3910471498966217, "learning_rate": 4.202316988007567e-05, "loss": 0.0253, "step": 11470 }, { "grad_norm": 0.4431130886077881, "learning_rate": 4.194156621146901e-05, "loss": 0.0216, "step": 11480 }, { "grad_norm": 0.4314492344856262, "learning_rate": 4.1859984574312596e-05, "loss": 0.0349, "step": 11490 }, { "grad_norm": 0.2151399701833725, "learning_rate": 4.177842519164752e-05, "loss": 0.0252, "step": 11500 }, { "grad_norm": 0.23166042566299438, "learning_rate": 4.169688828645404e-05, "loss": 0.0162, "step": 11510 }, { "grad_norm": 0.16609102487564087, "learning_rate": 4.161537408165092e-05, "loss": 0.0214, "step": 11520 }, { "grad_norm": 0.15997625887393951, "learning_rate": 4.1533882800094924e-05, "loss": 0.0231, "step": 11530 }, { "grad_norm": 0.21341855823993683, "learning_rate": 4.145241466458005e-05, "loss": 0.0276, "step": 11540 }, { "grad_norm": 0.42178070545196533, "learning_rate": 4.13709698978371e-05, "loss": 0.0251, "step": 11550 }, { "grad_norm": 0.18946197628974915, "learning_rate": 4.1289548722532944e-05, "loss": 0.0353, "step": 11560 }, { "grad_norm": 0.19865870475769043, "learning_rate": 4.120815136126999e-05, "loss": 0.0226, "step": 11570 }, { "grad_norm": 0.5400913953781128, "learning_rate": 4.112677803658548e-05, "loss": 0.0347, "step": 11580 }, { "grad_norm": 0.2605154812335968, "learning_rate": 4.1045428970951e-05, "loss": 0.0286, "step": 11590 }, { "grad_norm": 0.218913733959198, "learning_rate": 4.0964104386771785e-05, "loss": 0.0221, "step": 11600 }, { "grad_norm": 0.2642371356487274, "learning_rate": 4.0882804506386144e-05, "loss": 0.0474, "step": 11610 }, { "grad_norm": 0.4740236699581146, "learning_rate": 4.080152955206485e-05, "loss": 0.0375, "step": 11620 }, { "grad_norm": 0.29121097922325134, "learning_rate": 4.0720279746010505e-05, "loss": 0.048, "step": 11630 }, { "grad_norm": 0.2674061059951782, "learning_rate": 4.063905531035699e-05, "loss": 0.022, "step": 11640 }, { "grad_norm": 0.5309146642684937, "learning_rate": 4.055785646716882e-05, "loss": 0.0308, "step": 11650 }, { "grad_norm": 0.21866004168987274, "learning_rate": 4.047668343844051e-05, "loss": 0.0309, "step": 11660 }, { "grad_norm": 0.17855048179626465, "learning_rate": 4.039553644609604e-05, "loss": 0.0198, "step": 11670 }, { "grad_norm": 0.25463777780532837, "learning_rate": 4.0314415711988176e-05, "loss": 0.0288, "step": 11680 }, { "grad_norm": 0.332845538854599, "learning_rate": 4.023332145789792e-05, "loss": 0.0258, "step": 11690 }, { "grad_norm": 0.20202989876270294, "learning_rate": 4.015225390553385e-05, "loss": 0.0263, "step": 11700 }, { "grad_norm": 0.2870956361293793, "learning_rate": 4.007121327653158e-05, "loss": 0.026, "step": 11710 }, { "grad_norm": 0.3879694938659668, "learning_rate": 3.9990199792453064e-05, "loss": 0.0295, "step": 11720 }, { "grad_norm": 0.2177431285381317, "learning_rate": 3.9909213674786103e-05, "loss": 0.0206, "step": 11730 }, { "grad_norm": 0.3794606924057007, "learning_rate": 3.982825514494363e-05, "loss": 0.0334, "step": 11740 }, { "grad_norm": 0.1732175201177597, "learning_rate": 3.974732442426319e-05, "loss": 0.0299, "step": 11750 }, { "grad_norm": 0.20847928524017334, "learning_rate": 3.966642173400629e-05, "loss": 0.0294, "step": 11760 }, { "grad_norm": 0.44826966524124146, "learning_rate": 3.9585547295357764e-05, "loss": 0.0249, "step": 11770 }, { "grad_norm": 0.3483640253543854, "learning_rate": 3.950470132942526e-05, "loss": 0.022, "step": 11780 }, { "grad_norm": 0.3614063262939453, "learning_rate": 3.942388405723856e-05, "loss": 0.0312, "step": 11790 }, { "grad_norm": 0.2602802515029907, "learning_rate": 3.9343095699749e-05, "loss": 0.0225, "step": 11800 }, { "grad_norm": 0.21381863951683044, "learning_rate": 3.9262336477828874e-05, "loss": 0.0406, "step": 11810 }, { "grad_norm": 0.1822655200958252, "learning_rate": 3.9181606612270794e-05, "loss": 0.0227, "step": 11820 }, { "grad_norm": 0.36642977595329285, "learning_rate": 3.910090632378713e-05, "loss": 0.0366, "step": 11830 }, { "grad_norm": 0.20137616991996765, "learning_rate": 3.90202358330094e-05, "loss": 0.0204, "step": 11840 }, { "grad_norm": 0.20209522545337677, "learning_rate": 3.8939595360487656e-05, "loss": 0.0247, "step": 11850 }, { "grad_norm": 0.2902727425098419, "learning_rate": 3.885898512668984e-05, "loss": 0.0283, "step": 11860 }, { "grad_norm": 0.2715465724468231, "learning_rate": 3.877840535200127e-05, "loss": 0.0316, "step": 11870 }, { "grad_norm": 0.09178663790225983, "learning_rate": 3.869785625672397e-05, "loss": 0.0437, "step": 11880 }, { "grad_norm": 0.10313160717487335, "learning_rate": 3.8617338061076094e-05, "loss": 0.0251, "step": 11890 }, { "grad_norm": 0.3833531141281128, "learning_rate": 3.853685098519132e-05, "loss": 0.0352, "step": 11900 }, { "grad_norm": 0.3477344810962677, "learning_rate": 3.845639524911823e-05, "loss": 0.0334, "step": 11910 }, { "grad_norm": 0.3351706862449646, "learning_rate": 3.837597107281974e-05, "loss": 0.0204, "step": 11920 }, { "grad_norm": 0.6576680541038513, "learning_rate": 3.829557867617247e-05, "loss": 0.0311, "step": 11930 }, { "grad_norm": 0.27637162804603577, "learning_rate": 3.821521827896618e-05, "loss": 0.0235, "step": 11940 }, { "grad_norm": 0.18001775443553925, "learning_rate": 3.81348901009031e-05, "loss": 0.0233, "step": 11950 }, { "grad_norm": 0.30242615938186646, "learning_rate": 3.805459436159741e-05, "loss": 0.0308, "step": 11960 }, { "grad_norm": 0.3722730278968811, "learning_rate": 3.797433128057461e-05, "loss": 0.0362, "step": 11970 }, { "grad_norm": 0.3604329824447632, "learning_rate": 3.789410107727089e-05, "loss": 0.0258, "step": 11980 }, { "grad_norm": 0.12718746066093445, "learning_rate": 3.781390397103257e-05, "loss": 0.0263, "step": 11990 }, { "grad_norm": 0.405647337436676, "learning_rate": 3.7733740181115455e-05, "loss": 0.0271, "step": 12000 }, { "grad_norm": 0.34286415576934814, "learning_rate": 3.7653609926684306e-05, "loss": 0.0325, "step": 12010 }, { "grad_norm": 0.30191218852996826, "learning_rate": 3.757351342681217e-05, "loss": 0.0322, "step": 12020 }, { "grad_norm": 0.24962174892425537, "learning_rate": 3.749345090047982e-05, "loss": 0.0213, "step": 12030 }, { "grad_norm": 0.3531569838523865, "learning_rate": 3.741342256657515e-05, "loss": 0.0245, "step": 12040 }, { "grad_norm": 0.16351984441280365, "learning_rate": 3.7333428643892567e-05, "loss": 0.037, "step": 12050 }, { "grad_norm": 0.23649945855140686, "learning_rate": 3.725346935113239e-05, "loss": 0.0216, "step": 12060 }, { "grad_norm": 0.2519591450691223, "learning_rate": 3.717354490690029e-05, "loss": 0.0296, "step": 12070 }, { "grad_norm": 0.5084550976753235, "learning_rate": 3.709365552970664e-05, "loss": 0.0232, "step": 12080 }, { "grad_norm": 0.29651957750320435, "learning_rate": 3.7013801437965945e-05, "loss": 0.0166, "step": 12090 }, { "grad_norm": 0.42107775807380676, "learning_rate": 3.693398284999623e-05, "loss": 0.0318, "step": 12100 }, { "grad_norm": 0.30411913990974426, "learning_rate": 3.6854199984018484e-05, "loss": 0.0242, "step": 12110 }, { "grad_norm": 0.5815067887306213, "learning_rate": 3.677445305815601e-05, "loss": 0.0271, "step": 12120 }, { "grad_norm": 0.2332363724708557, "learning_rate": 3.669474229043387e-05, "loss": 0.0243, "step": 12130 }, { "grad_norm": 0.414297491312027, "learning_rate": 3.6615067898778235e-05, "loss": 0.0221, "step": 12140 }, { "grad_norm": 0.4023997485637665, "learning_rate": 3.6535430101015866e-05, "loss": 0.0274, "step": 12150 }, { "grad_norm": 0.2018202543258667, "learning_rate": 3.645582911487345e-05, "loss": 0.019, "step": 12160 }, { "grad_norm": 0.28000614047050476, "learning_rate": 3.637626515797706e-05, "loss": 0.0182, "step": 12170 }, { "grad_norm": 0.4380647838115692, "learning_rate": 3.629673844785152e-05, "loss": 0.029, "step": 12180 }, { "grad_norm": 0.4641306698322296, "learning_rate": 3.621724920191979e-05, "loss": 0.0252, "step": 12190 }, { "grad_norm": 0.2420637458562851, "learning_rate": 3.6137797637502444e-05, "loss": 0.042, "step": 12200 }, { "grad_norm": 0.2859507203102112, "learning_rate": 3.6058383971817035e-05, "loss": 0.0347, "step": 12210 }, { "grad_norm": 0.2930164933204651, "learning_rate": 3.59790084219775e-05, "loss": 0.0246, "step": 12220 }, { "grad_norm": 0.3332241177558899, "learning_rate": 3.589967120499353e-05, "loss": 0.0219, "step": 12230 }, { "grad_norm": 0.13815076649188995, "learning_rate": 3.5820372537770075e-05, "loss": 0.0196, "step": 12240 }, { "grad_norm": 0.3012155294418335, "learning_rate": 3.5741112637106655e-05, "loss": 0.0403, "step": 12250 }, { "grad_norm": 0.4452188313007355, "learning_rate": 3.5661891719696804e-05, "loss": 0.0326, "step": 12260 }, { "grad_norm": 0.23023128509521484, "learning_rate": 3.5582710002127504e-05, "loss": 0.0269, "step": 12270 }, { "grad_norm": 0.353941947221756, "learning_rate": 3.550356770087853e-05, "loss": 0.0207, "step": 12280 }, { "grad_norm": 0.42356541752815247, "learning_rate": 3.5424465032321914e-05, "loss": 0.0316, "step": 12290 }, { "grad_norm": 0.22531373798847198, "learning_rate": 3.5345402212721335e-05, "loss": 0.0271, "step": 12300 }, { "grad_norm": 0.3141672611236572, "learning_rate": 3.526637945823152e-05, "loss": 0.028, "step": 12310 }, { "grad_norm": 0.3621664345264435, "learning_rate": 3.518739698489767e-05, "loss": 0.0325, "step": 12320 }, { "grad_norm": 0.2679572105407715, "learning_rate": 3.510845500865485e-05, "loss": 0.0222, "step": 12330 }, { "grad_norm": 0.18188060820102692, "learning_rate": 3.502955374532739e-05, "loss": 0.0275, "step": 12340 }, { "grad_norm": 0.30384692549705505, "learning_rate": 3.495069341062836e-05, "loss": 0.0222, "step": 12350 }, { "grad_norm": 0.20437337458133698, "learning_rate": 3.4871874220158896e-05, "loss": 0.024, "step": 12360 }, { "grad_norm": 0.2626112103462219, "learning_rate": 3.479309638940762e-05, "loss": 0.0294, "step": 12370 }, { "grad_norm": 0.3420230746269226, "learning_rate": 3.4714360133750146e-05, "loss": 0.0212, "step": 12380 }, { "grad_norm": 0.23674432933330536, "learning_rate": 3.463566566844839e-05, "loss": 0.0345, "step": 12390 }, { "grad_norm": 0.310926616191864, "learning_rate": 3.4557013208650016e-05, "loss": 0.0284, "step": 12400 }, { "grad_norm": 0.5828701853752136, "learning_rate": 3.4478402969387857e-05, "loss": 0.0332, "step": 12410 }, { "grad_norm": 0.18565793335437775, "learning_rate": 3.4399835165579266e-05, "loss": 0.0201, "step": 12420 }, { "grad_norm": 0.48128461837768555, "learning_rate": 3.4321310012025645e-05, "loss": 0.0337, "step": 12430 }, { "grad_norm": 0.22701524198055267, "learning_rate": 3.424282772341176e-05, "loss": 0.0366, "step": 12440 }, { "grad_norm": 0.36069437861442566, "learning_rate": 3.416438851430519e-05, "loss": 0.0248, "step": 12450 }, { "grad_norm": 0.12960992753505707, "learning_rate": 3.408599259915577e-05, "loss": 0.0223, "step": 12460 }, { "grad_norm": 0.14093460142612457, "learning_rate": 3.400764019229487e-05, "loss": 0.022, "step": 12470 }, { "grad_norm": 0.30934223532676697, "learning_rate": 3.3929331507935035e-05, "loss": 0.0194, "step": 12480 }, { "grad_norm": 0.15594668686389923, "learning_rate": 3.3851066760169196e-05, "loss": 0.0189, "step": 12490 }, { "grad_norm": 0.23440571129322052, "learning_rate": 3.377284616297021e-05, "loss": 0.0274, "step": 12500 }, { "grad_norm": 0.3395257592201233, "learning_rate": 3.3694669930190166e-05, "loss": 0.0181, "step": 12510 }, { "grad_norm": 0.21954968571662903, "learning_rate": 3.36165382755599e-05, "loss": 0.0279, "step": 12520 }, { "grad_norm": 0.17136695981025696, "learning_rate": 3.35384514126884e-05, "loss": 0.0249, "step": 12530 }, { "grad_norm": 0.198324054479599, "learning_rate": 3.3460409555062154e-05, "loss": 0.0282, "step": 12540 }, { "grad_norm": 0.5399725437164307, "learning_rate": 3.3382412916044645e-05, "loss": 0.026, "step": 12550 }, { "grad_norm": 0.4857769310474396, "learning_rate": 3.330446170887566e-05, "loss": 0.0307, "step": 12560 }, { "grad_norm": 0.34405264258384705, "learning_rate": 3.3226556146670834e-05, "loss": 0.0235, "step": 12570 }, { "grad_norm": 0.20080843567848206, "learning_rate": 3.314869644242102e-05, "loss": 0.0257, "step": 12580 }, { "grad_norm": 0.38279613852500916, "learning_rate": 3.3070882808991674e-05, "loss": 0.0257, "step": 12590 }, { "grad_norm": 0.21805401146411896, "learning_rate": 3.2993115459122305e-05, "loss": 0.0222, "step": 12600 }, { "grad_norm": 0.2945617139339447, "learning_rate": 3.2915394605425835e-05, "loss": 0.0254, "step": 12610 }, { "grad_norm": 0.337274968624115, "learning_rate": 3.283772046038816e-05, "loss": 0.0254, "step": 12620 }, { "grad_norm": 0.3773317337036133, "learning_rate": 3.276009323636739e-05, "loss": 0.0191, "step": 12630 }, { "grad_norm": 0.41046056151390076, "learning_rate": 3.268251314559344e-05, "loss": 0.029, "step": 12640 }, { "grad_norm": 0.26641198992729187, "learning_rate": 3.2604980400167254e-05, "loss": 0.0261, "step": 12650 }, { "grad_norm": 0.39587777853012085, "learning_rate": 3.252749521206042e-05, "loss": 0.0424, "step": 12660 }, { "grad_norm": 0.25540727376937866, "learning_rate": 3.2450057793114494e-05, "loss": 0.0237, "step": 12670 }, { "grad_norm": 0.17594082653522491, "learning_rate": 3.2372668355040435e-05, "loss": 0.0317, "step": 12680 }, { "grad_norm": 0.6507959365844727, "learning_rate": 3.2295327109418005e-05, "loss": 0.025, "step": 12690 }, { "grad_norm": 0.24934737384319305, "learning_rate": 3.221803426769518e-05, "loss": 0.0366, "step": 12700 }, { "grad_norm": 0.2510274350643158, "learning_rate": 3.214079004118768e-05, "loss": 0.0197, "step": 12710 }, { "grad_norm": 0.3049881160259247, "learning_rate": 3.2063594641078234e-05, "loss": 0.0217, "step": 12720 }, { "grad_norm": 0.35464248061180115, "learning_rate": 3.198644827841616e-05, "loss": 0.0171, "step": 12730 }, { "grad_norm": 0.3850623071193695, "learning_rate": 3.1909351164116654e-05, "loss": 0.0256, "step": 12740 }, { "grad_norm": 0.18634547293186188, "learning_rate": 3.183230350896026e-05, "loss": 0.0237, "step": 12750 }, { "grad_norm": 0.5240119099617004, "learning_rate": 3.1755305523592337e-05, "loss": 0.0259, "step": 12760 }, { "grad_norm": 0.275981605052948, "learning_rate": 3.167835741852245e-05, "loss": 0.0277, "step": 12770 }, { "grad_norm": 0.2884126305580139, "learning_rate": 3.160145940412378e-05, "loss": 0.0311, "step": 12780 }, { "grad_norm": 0.31567177176475525, "learning_rate": 3.1524611690632545e-05, "loss": 0.024, "step": 12790 }, { "grad_norm": 0.2496463805437088, "learning_rate": 3.144781448814746e-05, "loss": 0.0232, "step": 12800 }, { "grad_norm": 0.2244059443473816, "learning_rate": 3.1371068006629145e-05, "loss": 0.0233, "step": 12810 }, { "grad_norm": 0.26163628697395325, "learning_rate": 3.129437245589956e-05, "loss": 0.0296, "step": 12820 }, { "grad_norm": 0.4849525988101959, "learning_rate": 3.121772804564143e-05, "loss": 0.0279, "step": 12830 }, { "grad_norm": 0.21702219545841217, "learning_rate": 3.11411349853976e-05, "loss": 0.027, "step": 12840 }, { "grad_norm": 0.2369028627872467, "learning_rate": 3.10645934845706e-05, "loss": 0.0187, "step": 12850 }, { "grad_norm": 0.35298478603363037, "learning_rate": 3.098810375242196e-05, "loss": 0.0299, "step": 12860 }, { "grad_norm": 0.17053690552711487, "learning_rate": 3.0911665998071704e-05, "loss": 0.0183, "step": 12870 }, { "grad_norm": 0.21893882751464844, "learning_rate": 3.083528043049774e-05, "loss": 0.0231, "step": 12880 }, { "grad_norm": 0.1868700236082077, "learning_rate": 3.0758947258535255e-05, "loss": 0.0209, "step": 12890 }, { "grad_norm": 0.30209845304489136, "learning_rate": 3.068266669087625e-05, "loss": 0.0217, "step": 12900 }, { "grad_norm": 0.22564226388931274, "learning_rate": 3.060643893606887e-05, "loss": 0.0206, "step": 12910 }, { "grad_norm": 0.3218061029911041, "learning_rate": 3.053026420251693e-05, "loss": 0.0114, "step": 12920 }, { "grad_norm": 0.16161943972110748, "learning_rate": 3.0454142698479183e-05, "loss": 0.0238, "step": 12930 }, { "grad_norm": 0.39317476749420166, "learning_rate": 3.0378074632068954e-05, "loss": 0.0195, "step": 12940 }, { "grad_norm": 0.12611640989780426, "learning_rate": 3.0302060211253408e-05, "loss": 0.016, "step": 12950 }, { "grad_norm": 0.3329657316207886, "learning_rate": 3.0226099643853073e-05, "loss": 0.0229, "step": 12960 }, { "grad_norm": 0.32624730467796326, "learning_rate": 3.0150193137541283e-05, "loss": 0.0186, "step": 12970 }, { "grad_norm": 0.24989771842956543, "learning_rate": 3.0074340899843467e-05, "loss": 0.0192, "step": 12980 }, { "grad_norm": 0.10836402326822281, "learning_rate": 2.999854313813677e-05, "loss": 0.0246, "step": 12990 }, { "grad_norm": 0.3700701892375946, "learning_rate": 2.9922800059649382e-05, "loss": 0.0241, "step": 13000 }, { "grad_norm": 0.3759732246398926, "learning_rate": 2.9847111871459976e-05, "loss": 0.0357, "step": 13010 }, { "grad_norm": 0.3690287172794342, "learning_rate": 2.977147878049721e-05, "loss": 0.0268, "step": 13020 }, { "grad_norm": 0.2705408036708832, "learning_rate": 2.9695900993539006e-05, "loss": 0.0258, "step": 13030 }, { "grad_norm": 0.27386996150016785, "learning_rate": 2.9620378717212183e-05, "loss": 0.0232, "step": 13040 }, { "grad_norm": 0.2577216923236847, "learning_rate": 2.9544912157991745e-05, "loss": 0.0176, "step": 13050 }, { "grad_norm": 0.27413344383239746, "learning_rate": 2.9469501522200405e-05, "loss": 0.026, "step": 13060 }, { "grad_norm": 0.1716402769088745, "learning_rate": 2.9394147016007946e-05, "loss": 0.027, "step": 13070 }, { "grad_norm": 0.20654554665088654, "learning_rate": 2.9318848845430702e-05, "loss": 0.0238, "step": 13080 }, { "grad_norm": 0.2717389166355133, "learning_rate": 2.9243607216331013e-05, "loss": 0.0289, "step": 13090 }, { "grad_norm": 0.18845367431640625, "learning_rate": 2.916842233441661e-05, "loss": 0.0245, "step": 13100 }, { "grad_norm": 0.22851665318012238, "learning_rate": 2.90932944052401e-05, "loss": 0.0205, "step": 13110 }, { "grad_norm": 0.10635247826576233, "learning_rate": 2.9018223634198354e-05, "loss": 0.0196, "step": 13120 }, { "grad_norm": 0.3600257635116577, "learning_rate": 2.8943210226532025e-05, "loss": 0.0342, "step": 13130 }, { "grad_norm": 0.4032753109931946, "learning_rate": 2.8868254387324857e-05, "loss": 0.0252, "step": 13140 }, { "grad_norm": 0.5242461562156677, "learning_rate": 2.8793356321503306e-05, "loss": 0.0281, "step": 13150 }, { "grad_norm": 0.17106902599334717, "learning_rate": 2.87185162338358e-05, "loss": 0.0201, "step": 13160 }, { "grad_norm": 0.22562071681022644, "learning_rate": 2.8643734328932253e-05, "loss": 0.0198, "step": 13170 }, { "grad_norm": 0.13923519849777222, "learning_rate": 2.856901081124359e-05, "loss": 0.0217, "step": 13180 }, { "grad_norm": 0.2287658452987671, "learning_rate": 2.8494345885061002e-05, "loss": 0.0196, "step": 13190 }, { "grad_norm": 0.5717466473579407, "learning_rate": 2.8419739754515616e-05, "loss": 0.022, "step": 13200 }, { "grad_norm": 0.11969102174043655, "learning_rate": 2.8345192623577666e-05, "loss": 0.0157, "step": 13210 }, { "grad_norm": 0.21323078870773315, "learning_rate": 2.8270704696056193e-05, "loss": 0.0233, "step": 13220 }, { "grad_norm": 0.36979830265045166, "learning_rate": 2.8196276175598367e-05, "loss": 0.0317, "step": 13230 }, { "grad_norm": 0.2436993420124054, "learning_rate": 2.8121907265688884e-05, "loss": 0.0231, "step": 13240 }, { "grad_norm": 0.2477279007434845, "learning_rate": 2.804759816964957e-05, "loss": 0.0164, "step": 13250 }, { "grad_norm": 0.5008840560913086, "learning_rate": 2.797334909063857e-05, "loss": 0.015, "step": 13260 }, { "grad_norm": 0.11843416839838028, "learning_rate": 2.7899160231650056e-05, "loss": 0.02, "step": 13270 }, { "grad_norm": 0.20259647071361542, "learning_rate": 2.7825031795513585e-05, "loss": 0.0192, "step": 13280 }, { "grad_norm": 0.2110484540462494, "learning_rate": 2.775096398489341e-05, "loss": 0.0216, "step": 13290 }, { "grad_norm": 0.2280147224664688, "learning_rate": 2.7676957002288163e-05, "loss": 0.0204, "step": 13300 }, { "grad_norm": 0.19791455566883087, "learning_rate": 2.760301105003003e-05, "loss": 0.0308, "step": 13310 }, { "grad_norm": 0.6110244989395142, "learning_rate": 2.752912633028446e-05, "loss": 0.0339, "step": 13320 }, { "grad_norm": 0.3705899119377136, "learning_rate": 2.7455303045049474e-05, "loss": 0.0348, "step": 13330 }, { "grad_norm": 0.14164911210536957, "learning_rate": 2.7381541396155098e-05, "loss": 0.0195, "step": 13340 }, { "grad_norm": 0.20682582259178162, "learning_rate": 2.730784158526286e-05, "loss": 0.0205, "step": 13350 }, { "grad_norm": 0.10009776800870895, "learning_rate": 2.723420381386521e-05, "loss": 0.024, "step": 13360 }, { "grad_norm": 0.17733310163021088, "learning_rate": 2.7160628283285018e-05, "loss": 0.0267, "step": 13370 }, { "grad_norm": 0.543793261051178, "learning_rate": 2.7087115194675007e-05, "loss": 0.029, "step": 13380 }, { "grad_norm": 0.26760250329971313, "learning_rate": 2.701366474901712e-05, "loss": 0.0295, "step": 13390 }, { "grad_norm": 0.1812628209590912, "learning_rate": 2.6940277147122085e-05, "loss": 0.0211, "step": 13400 }, { "grad_norm": 0.15894292294979095, "learning_rate": 2.686695258962878e-05, "loss": 0.0202, "step": 13410 }, { "grad_norm": 0.12431477010250092, "learning_rate": 2.679369127700375e-05, "loss": 0.0273, "step": 13420 }, { "grad_norm": 0.407216340303421, "learning_rate": 2.672049340954067e-05, "loss": 0.0186, "step": 13430 }, { "grad_norm": 0.2614075541496277, "learning_rate": 2.6647359187359676e-05, "loss": 0.0288, "step": 13440 }, { "grad_norm": 0.4476523697376251, "learning_rate": 2.6574288810406946e-05, "loss": 0.0315, "step": 13450 }, { "grad_norm": 0.19416925311088562, "learning_rate": 2.6501282478454083e-05, "loss": 0.0286, "step": 13460 }, { "grad_norm": 1.1362308263778687, "learning_rate": 2.6428340391097618e-05, "loss": 0.0306, "step": 13470 }, { "grad_norm": 0.44146788120269775, "learning_rate": 2.6355462747758485e-05, "loss": 0.0382, "step": 13480 }, { "grad_norm": 0.20890440046787262, "learning_rate": 2.6282649747681304e-05, "loss": 0.0213, "step": 13490 }, { "grad_norm": 0.31235355138778687, "learning_rate": 2.620990158993406e-05, "loss": 0.0125, "step": 13500 }, { "grad_norm": 0.32159560918807983, "learning_rate": 2.6137218473407477e-05, "loss": 0.026, "step": 13510 }, { "grad_norm": 0.16577616333961487, "learning_rate": 2.606460059681436e-05, "loss": 0.0346, "step": 13520 }, { "grad_norm": 0.429965078830719, "learning_rate": 2.599204815868928e-05, "loss": 0.0381, "step": 13530 }, { "grad_norm": 0.17882920801639557, "learning_rate": 2.5919561357387756e-05, "loss": 0.0237, "step": 13540 }, { "grad_norm": 0.2290530800819397, "learning_rate": 2.5847140391085972e-05, "loss": 0.0201, "step": 13550 }, { "grad_norm": 0.20041543245315552, "learning_rate": 2.5774785457780103e-05, "loss": 0.0324, "step": 13560 }, { "grad_norm": 0.1382402628660202, "learning_rate": 2.5702496755285753e-05, "loss": 0.0226, "step": 13570 }, { "grad_norm": 0.4306432008743286, "learning_rate": 2.5630274481237483e-05, "loss": 0.0204, "step": 13580 }, { "grad_norm": 0.12714605033397675, "learning_rate": 2.5558118833088197e-05, "loss": 0.0351, "step": 13590 }, { "grad_norm": 0.38723865151405334, "learning_rate": 2.548603000810872e-05, "loss": 0.0421, "step": 13600 }, { "grad_norm": 0.234077587723732, "learning_rate": 2.5414008203387152e-05, "loss": 0.02, "step": 13610 }, { "grad_norm": 0.20112663507461548, "learning_rate": 2.534205361582834e-05, "loss": 0.037, "step": 13620 }, { "grad_norm": 0.243337482213974, "learning_rate": 2.527016644215338e-05, "loss": 0.0218, "step": 13630 }, { "grad_norm": 0.3914124369621277, "learning_rate": 2.519834687889905e-05, "loss": 0.0246, "step": 13640 }, { "grad_norm": 0.4395959675312042, "learning_rate": 2.5126595122417295e-05, "loss": 0.0267, "step": 13650 }, { "grad_norm": 0.576765775680542, "learning_rate": 2.5054911368874713e-05, "loss": 0.0246, "step": 13660 }, { "grad_norm": 0.20656850934028625, "learning_rate": 2.4983295814251916e-05, "loss": 0.0233, "step": 13670 }, { "grad_norm": 0.12873475253582, "learning_rate": 2.4911748654343105e-05, "loss": 0.02, "step": 13680 }, { "grad_norm": 0.5158886909484863, "learning_rate": 2.4840270084755463e-05, "loss": 0.0226, "step": 13690 }, { "grad_norm": 0.32906392216682434, "learning_rate": 2.4768860300908685e-05, "loss": 0.0174, "step": 13700 }, { "grad_norm": 0.21135656535625458, "learning_rate": 2.469751949803443e-05, "loss": 0.0228, "step": 13710 }, { "grad_norm": 0.1754065603017807, "learning_rate": 2.4626247871175666e-05, "loss": 0.0181, "step": 13720 }, { "grad_norm": 0.18912409245967865, "learning_rate": 2.4555045615186346e-05, "loss": 0.017, "step": 13730 }, { "grad_norm": 0.15924926102161407, "learning_rate": 2.4483912924730677e-05, "loss": 0.02, "step": 13740 }, { "grad_norm": 0.31903016567230225, "learning_rate": 2.4412849994282742e-05, "loss": 0.019, "step": 13750 }, { "grad_norm": 0.3755442500114441, "learning_rate": 2.434185701812592e-05, "loss": 0.0291, "step": 13760 }, { "grad_norm": 0.2624180316925049, "learning_rate": 2.4270934190352218e-05, "loss": 0.0151, "step": 13770 }, { "grad_norm": 0.3451279401779175, "learning_rate": 2.4200081704861998e-05, "loss": 0.0183, "step": 13780 }, { "grad_norm": 0.4145757555961609, "learning_rate": 2.412929975536321e-05, "loss": 0.0218, "step": 13790 }, { "grad_norm": 0.3473895490169525, "learning_rate": 2.4058588535371017e-05, "loss": 0.016, "step": 13800 }, { "grad_norm": 0.1519661694765091, "learning_rate": 2.3987948238207243e-05, "loss": 0.0354, "step": 13810 }, { "grad_norm": 0.17228980362415314, "learning_rate": 2.3917379056999678e-05, "loss": 0.0258, "step": 13820 }, { "grad_norm": 0.24880445003509521, "learning_rate": 2.3846881184681824e-05, "loss": 0.0252, "step": 13830 }, { "grad_norm": 0.22821317613124847, "learning_rate": 2.377645481399214e-05, "loss": 0.0332, "step": 13840 }, { "grad_norm": 0.3902832567691803, "learning_rate": 2.3706100137473667e-05, "loss": 0.025, "step": 13850 }, { "grad_norm": 0.1672590970993042, "learning_rate": 2.3635817347473394e-05, "loss": 0.0276, "step": 13860 }, { "grad_norm": 0.17667557299137115, "learning_rate": 2.3565606636141757e-05, "loss": 0.021, "step": 13870 }, { "grad_norm": 0.19600187242031097, "learning_rate": 2.3495468195432203e-05, "loss": 0.0177, "step": 13880 }, { "grad_norm": 0.31857743859291077, "learning_rate": 2.3425402217100507e-05, "loss": 0.0257, "step": 13890 }, { "grad_norm": 0.21914225816726685, "learning_rate": 2.3355408892704424e-05, "loss": 0.0181, "step": 13900 }, { "grad_norm": 0.3824993968009949, "learning_rate": 2.3285488413603003e-05, "loss": 0.0218, "step": 13910 }, { "grad_norm": 0.4360935091972351, "learning_rate": 2.321564097095615e-05, "loss": 0.0239, "step": 13920 }, { "grad_norm": 0.18124280869960785, "learning_rate": 2.3145866755724142e-05, "loss": 0.0202, "step": 13930 }, { "grad_norm": 0.3161773681640625, "learning_rate": 2.307616595866699e-05, "loss": 0.0161, "step": 13940 }, { "grad_norm": 0.23667371273040771, "learning_rate": 2.3006538770344032e-05, "loss": 0.0205, "step": 13950 }, { "grad_norm": 0.2887951731681824, "learning_rate": 2.293698538111334e-05, "loss": 0.0206, "step": 13960 }, { "grad_norm": 0.21399231255054474, "learning_rate": 2.28675059811312e-05, "loss": 0.0327, "step": 13970 }, { "grad_norm": 0.1776636391878128, "learning_rate": 2.279810076035167e-05, "loss": 0.0286, "step": 13980 }, { "grad_norm": 0.6997559070587158, "learning_rate": 2.272876990852596e-05, "loss": 0.0247, "step": 13990 }, { "grad_norm": 0.14329971373081207, "learning_rate": 2.265951361520195e-05, "loss": 0.0166, "step": 14000 }, { "grad_norm": 0.1624142974615097, "learning_rate": 2.2590332069723748e-05, "loss": 0.0208, "step": 14010 }, { "grad_norm": 0.28449195623397827, "learning_rate": 2.2521225461231004e-05, "loss": 0.0308, "step": 14020 }, { "grad_norm": 0.4659704267978668, "learning_rate": 2.2452193978658597e-05, "loss": 0.0309, "step": 14030 }, { "grad_norm": 0.3407534658908844, "learning_rate": 2.238323781073594e-05, "loss": 0.0228, "step": 14040 }, { "grad_norm": 0.4123205840587616, "learning_rate": 2.2314357145986552e-05, "loss": 0.0261, "step": 14050 }, { "grad_norm": 0.172753244638443, "learning_rate": 2.224555217272757e-05, "loss": 0.029, "step": 14060 }, { "grad_norm": 0.4571063220500946, "learning_rate": 2.2176823079069127e-05, "loss": 0.0264, "step": 14070 }, { "grad_norm": 0.1457226276397705, "learning_rate": 2.210817005291398e-05, "loss": 0.0135, "step": 14080 }, { "grad_norm": 0.21855150163173676, "learning_rate": 2.203959328195686e-05, "loss": 0.0165, "step": 14090 }, { "grad_norm": 0.1833777129650116, "learning_rate": 2.1971092953684026e-05, "loss": 0.0218, "step": 14100 }, { "grad_norm": 0.20567305386066437, "learning_rate": 2.1902669255372788e-05, "loss": 0.0202, "step": 14110 }, { "grad_norm": 0.15128324925899506, "learning_rate": 2.1834322374090897e-05, "loss": 0.0236, "step": 14120 }, { "grad_norm": 0.3515031337738037, "learning_rate": 2.1766052496696153e-05, "loss": 0.0283, "step": 14130 }, { "grad_norm": 0.39070484042167664, "learning_rate": 2.169785980983577e-05, "loss": 0.0303, "step": 14140 }, { "grad_norm": 0.47723621129989624, "learning_rate": 2.162974449994593e-05, "loss": 0.0181, "step": 14150 }, { "grad_norm": 0.4971657693386078, "learning_rate": 2.1561706753251337e-05, "loss": 0.0403, "step": 14160 }, { "grad_norm": 0.3704018294811249, "learning_rate": 2.1493746755764544e-05, "loss": 0.0223, "step": 14170 }, { "grad_norm": 0.13312096893787384, "learning_rate": 2.1425864693285635e-05, "loss": 0.0219, "step": 14180 }, { "grad_norm": 0.4792286455631256, "learning_rate": 2.1358060751401547e-05, "loss": 0.0253, "step": 14190 }, { "grad_norm": 0.31885582208633423, "learning_rate": 2.129033511548566e-05, "loss": 0.0221, "step": 14200 }, { "grad_norm": 0.11596840620040894, "learning_rate": 2.1222687970697315e-05, "loss": 0.0243, "step": 14210 }, { "grad_norm": 0.32194647192955017, "learning_rate": 2.1155119501981173e-05, "loss": 0.0271, "step": 14220 }, { "grad_norm": 0.35104992985725403, "learning_rate": 2.1087629894066895e-05, "loss": 0.0164, "step": 14230 }, { "grad_norm": 0.6368436813354492, "learning_rate": 2.1020219331468473e-05, "loss": 0.0232, "step": 14240 }, { "grad_norm": 0.36711663007736206, "learning_rate": 2.095288799848379e-05, "loss": 0.0235, "step": 14250 }, { "grad_norm": 0.2038203477859497, "learning_rate": 2.088563607919417e-05, "loss": 0.027, "step": 14260 }, { "grad_norm": 0.26954466104507446, "learning_rate": 2.0818463757463786e-05, "loss": 0.0291, "step": 14270 }, { "grad_norm": 0.45064520835876465, "learning_rate": 2.0751371216939175e-05, "loss": 0.0227, "step": 14280 }, { "grad_norm": 0.2357233762741089, "learning_rate": 2.068435864104882e-05, "loss": 0.0187, "step": 14290 }, { "grad_norm": 0.23728369176387787, "learning_rate": 2.0617426213002506e-05, "loss": 0.0267, "step": 14300 }, { "grad_norm": 0.2995329201221466, "learning_rate": 2.055057411579097e-05, "loss": 0.0179, "step": 14310 }, { "grad_norm": 0.19889546930789948, "learning_rate": 2.0483802532185286e-05, "loss": 0.0258, "step": 14320 }, { "grad_norm": 0.2262890636920929, "learning_rate": 2.041711164473638e-05, "loss": 0.019, "step": 14330 }, { "grad_norm": 0.2174612581729889, "learning_rate": 2.0350501635774637e-05, "loss": 0.0207, "step": 14340 }, { "grad_norm": 0.8958914279937744, "learning_rate": 2.0283972687409247e-05, "loss": 0.0253, "step": 14350 }, { "grad_norm": 0.36986058950424194, "learning_rate": 2.021752498152784e-05, "loss": 0.0222, "step": 14360 }, { "grad_norm": 0.2585403323173523, "learning_rate": 2.015115869979589e-05, "loss": 0.0249, "step": 14370 }, { "grad_norm": 0.35662907361984253, "learning_rate": 2.0084874023656265e-05, "loss": 0.0216, "step": 14380 }, { "grad_norm": 0.36216458678245544, "learning_rate": 2.001867113432877e-05, "loss": 0.0219, "step": 14390 }, { "grad_norm": 0.3488549292087555, "learning_rate": 1.995255021280954e-05, "loss": 0.0186, "step": 14400 }, { "grad_norm": 0.28035154938697815, "learning_rate": 1.9886511439870688e-05, "loss": 0.018, "step": 14410 }, { "grad_norm": 0.17905603349208832, "learning_rate": 1.9820554996059675e-05, "loss": 0.0278, "step": 14420 }, { "grad_norm": 0.17527271807193756, "learning_rate": 1.9754681061698893e-05, "loss": 0.0199, "step": 14430 }, { "grad_norm": 0.24398522078990936, "learning_rate": 1.9688889816885185e-05, "loss": 0.0177, "step": 14440 }, { "grad_norm": 0.22834695875644684, "learning_rate": 1.962318144148928e-05, "loss": 0.0207, "step": 14450 }, { "grad_norm": 0.20363375544548035, "learning_rate": 1.955755611515539e-05, "loss": 0.0305, "step": 14460 }, { "grad_norm": 0.16525743901729584, "learning_rate": 1.9492014017300642e-05, "loss": 0.0249, "step": 14470 }, { "grad_norm": 0.23352274298667908, "learning_rate": 1.942655532711461e-05, "loss": 0.0149, "step": 14480 }, { "grad_norm": 0.4155940115451813, "learning_rate": 1.9361180223558882e-05, "loss": 0.0172, "step": 14490 }, { "grad_norm": 0.14329977333545685, "learning_rate": 1.929588888536647e-05, "loss": 0.0273, "step": 14500 }, { "grad_norm": 0.163722962141037, "learning_rate": 1.9230681491041425e-05, "loss": 0.023, "step": 14510 }, { "grad_norm": 0.2293214350938797, "learning_rate": 1.9165558218858264e-05, "loss": 0.019, "step": 14520 }, { "grad_norm": 0.25876879692077637, "learning_rate": 1.9100519246861505e-05, "loss": 0.0309, "step": 14530 }, { "grad_norm": 0.12151789665222168, "learning_rate": 1.9035564752865248e-05, "loss": 0.0163, "step": 14540 }, { "grad_norm": 0.28978556394577026, "learning_rate": 1.897069491445258e-05, "loss": 0.0241, "step": 14550 }, { "grad_norm": 0.25811681151390076, "learning_rate": 1.890590990897515e-05, "loss": 0.0205, "step": 14560 }, { "grad_norm": 0.17693071067333221, "learning_rate": 1.884120991355272e-05, "loss": 0.0207, "step": 14570 }, { "grad_norm": 0.47872528433799744, "learning_rate": 1.8776595105072576e-05, "loss": 0.0277, "step": 14580 }, { "grad_norm": 0.1223321482539177, "learning_rate": 1.8712065660189166e-05, "loss": 0.0372, "step": 14590 }, { "grad_norm": 0.14384602010250092, "learning_rate": 1.8647621755323513e-05, "loss": 0.0298, "step": 14600 }, { "grad_norm": 0.08701182156801224, "learning_rate": 1.858326356666278e-05, "loss": 0.0226, "step": 14610 }, { "grad_norm": 0.23272404074668884, "learning_rate": 1.851899127015983e-05, "loss": 0.0136, "step": 14620 }, { "grad_norm": 0.5119808912277222, "learning_rate": 1.8454805041532626e-05, "loss": 0.0133, "step": 14630 }, { "grad_norm": 0.1447504609823227, "learning_rate": 1.8390705056263906e-05, "loss": 0.0138, "step": 14640 }, { "grad_norm": 0.3530803620815277, "learning_rate": 1.832669148960057e-05, "loss": 0.0297, "step": 14650 }, { "grad_norm": 0.1668897122144699, "learning_rate": 1.8262764516553233e-05, "loss": 0.0168, "step": 14660 }, { "grad_norm": 0.2017683982849121, "learning_rate": 1.8198924311895843e-05, "loss": 0.0268, "step": 14670 }, { "grad_norm": 0.22527383267879486, "learning_rate": 1.813517105016505e-05, "loss": 0.0177, "step": 14680 }, { "grad_norm": 0.2914533317089081, "learning_rate": 1.8071504905659888e-05, "loss": 0.0226, "step": 14690 }, { "grad_norm": 0.37014755606651306, "learning_rate": 1.800792605244109e-05, "loss": 0.0198, "step": 14700 }, { "grad_norm": 0.42177221179008484, "learning_rate": 1.7944434664330844e-05, "loss": 0.0187, "step": 14710 }, { "grad_norm": 0.10951755940914154, "learning_rate": 1.7881030914912212e-05, "loss": 0.0235, "step": 14720 }, { "grad_norm": 0.36568406224250793, "learning_rate": 1.7817714977528577e-05, "loss": 0.0214, "step": 14730 }, { "grad_norm": 0.14736436307430267, "learning_rate": 1.7754487025283332e-05, "loss": 0.0241, "step": 14740 }, { "grad_norm": 0.38588741421699524, "learning_rate": 1.7691347231039275e-05, "loss": 0.0252, "step": 14750 }, { "grad_norm": 0.37550488114356995, "learning_rate": 1.7628295767418164e-05, "loss": 0.0197, "step": 14760 }, { "grad_norm": 0.3674973249435425, "learning_rate": 1.7565332806800333e-05, "loss": 0.019, "step": 14770 }, { "grad_norm": 0.4130190908908844, "learning_rate": 1.750245852132408e-05, "loss": 0.0257, "step": 14780 }, { "grad_norm": 0.2807732820510864, "learning_rate": 1.7439673082885323e-05, "loss": 0.0186, "step": 14790 }, { "grad_norm": 0.3884442150592804, "learning_rate": 1.7376976663137047e-05, "loss": 0.0154, "step": 14800 }, { "grad_norm": 0.2679819166660309, "learning_rate": 1.7314369433488853e-05, "loss": 0.0207, "step": 14810 }, { "grad_norm": 0.21836309134960175, "learning_rate": 1.7251851565106548e-05, "loss": 0.0241, "step": 14820 }, { "grad_norm": 0.17943395674228668, "learning_rate": 1.7189423228911574e-05, "loss": 0.0259, "step": 14830 }, { "grad_norm": 0.13569554686546326, "learning_rate": 1.7127084595580606e-05, "loss": 0.0322, "step": 14840 }, { "grad_norm": 0.15200182795524597, "learning_rate": 1.706483583554513e-05, "loss": 0.0243, "step": 14850 }, { "grad_norm": 0.28119462728500366, "learning_rate": 1.700267711899083e-05, "loss": 0.0197, "step": 14860 }, { "grad_norm": 0.2913888692855835, "learning_rate": 1.69406086158573e-05, "loss": 0.017, "step": 14870 }, { "grad_norm": 0.2842000424861908, "learning_rate": 1.6878630495837455e-05, "loss": 0.0263, "step": 14880 }, { "grad_norm": 0.36040598154067993, "learning_rate": 1.681674292837707e-05, "loss": 0.0167, "step": 14890 }, { "grad_norm": 0.20902203023433685, "learning_rate": 1.6754946082674444e-05, "loss": 0.0146, "step": 14900 }, { "grad_norm": 0.40076181292533875, "learning_rate": 1.6693240127679748e-05, "loss": 0.0172, "step": 14910 }, { "grad_norm": 0.25875675678253174, "learning_rate": 1.663162523209475e-05, "loss": 0.0179, "step": 14920 }, { "grad_norm": 0.43915289640426636, "learning_rate": 1.6570101564372193e-05, "loss": 0.0197, "step": 14930 }, { "grad_norm": 0.13455311954021454, "learning_rate": 1.650866929271543e-05, "loss": 0.0124, "step": 14940 }, { "grad_norm": 0.36437922716140747, "learning_rate": 1.644732858507797e-05, "loss": 0.0234, "step": 14950 }, { "grad_norm": 0.07522600889205933, "learning_rate": 1.6386079609162943e-05, "loss": 0.0194, "step": 14960 }, { "grad_norm": 0.14662720263004303, "learning_rate": 1.6324922532422742e-05, "loss": 0.0259, "step": 14970 }, { "grad_norm": 0.16744199395179749, "learning_rate": 1.6263857522058434e-05, "loss": 0.0213, "step": 14980 }, { "grad_norm": 0.14888857305049896, "learning_rate": 1.6202884745019443e-05, "loss": 0.0143, "step": 14990 }, { "grad_norm": 0.1907859593629837, "learning_rate": 1.614200436800304e-05, "loss": 0.0261, "step": 15000 }, { "grad_norm": 0.08187828212976456, "learning_rate": 1.6081216557453814e-05, "loss": 0.0204, "step": 15010 }, { "grad_norm": 0.30506542325019836, "learning_rate": 1.6020521479563367e-05, "loss": 0.0159, "step": 15020 }, { "grad_norm": 0.3486149311065674, "learning_rate": 1.5959919300269654e-05, "loss": 0.0299, "step": 15030 }, { "grad_norm": 0.2508726716041565, "learning_rate": 1.5899410185256764e-05, "loss": 0.0279, "step": 15040 }, { "grad_norm": 0.328937292098999, "learning_rate": 1.583899429995431e-05, "loss": 0.0237, "step": 15050 }, { "grad_norm": 0.2715276777744293, "learning_rate": 1.5778671809536993e-05, "loss": 0.0219, "step": 15060 }, { "grad_norm": 0.20292209088802338, "learning_rate": 1.5718442878924246e-05, "loss": 0.0274, "step": 15070 }, { "grad_norm": 0.13971933722496033, "learning_rate": 1.5658307672779593e-05, "loss": 0.0184, "step": 15080 }, { "grad_norm": 0.47437334060668945, "learning_rate": 1.5598266355510427e-05, "loss": 0.0287, "step": 15090 }, { "grad_norm": 0.2329227775335312, "learning_rate": 1.553831909126744e-05, "loss": 0.0179, "step": 15100 }, { "grad_norm": 0.1932559758424759, "learning_rate": 1.5478466043944135e-05, "loss": 0.0114, "step": 15110 }, { "grad_norm": 0.10852614045143127, "learning_rate": 1.5418707377176468e-05, "loss": 0.0145, "step": 15120 }, { "grad_norm": 0.16655923426151276, "learning_rate": 1.535904325434233e-05, "loss": 0.0204, "step": 15130 }, { "grad_norm": 0.2825092375278473, "learning_rate": 1.529947383856118e-05, "loss": 0.0157, "step": 15140 }, { "grad_norm": 0.18802982568740845, "learning_rate": 1.5239999292693524e-05, "loss": 0.0263, "step": 15150 }, { "grad_norm": 0.27893686294555664, "learning_rate": 1.5180619779340505e-05, "loss": 0.0307, "step": 15160 }, { "grad_norm": 0.30747687816619873, "learning_rate": 1.5121335460843428e-05, "loss": 0.0263, "step": 15170 }, { "grad_norm": 0.20702841877937317, "learning_rate": 1.5062146499283347e-05, "loss": 0.0168, "step": 15180 }, { "grad_norm": 0.21456898748874664, "learning_rate": 1.5003053056480643e-05, "loss": 0.0154, "step": 15190 }, { "grad_norm": 0.3981521725654602, "learning_rate": 1.4944055293994551e-05, "loss": 0.0267, "step": 15200 }, { "grad_norm": 0.09997072070837021, "learning_rate": 1.4885153373122656e-05, "loss": 0.021, "step": 15210 }, { "grad_norm": 0.4273131489753723, "learning_rate": 1.482634745490059e-05, "loss": 0.0156, "step": 15220 }, { "grad_norm": 0.09821368753910065, "learning_rate": 1.4767637700101466e-05, "loss": 0.0196, "step": 15230 }, { "grad_norm": 0.13168549537658691, "learning_rate": 1.4709024269235528e-05, "loss": 0.016, "step": 15240 }, { "grad_norm": 0.3125094473361969, "learning_rate": 1.4650507322549684e-05, "loss": 0.03, "step": 15250 }, { "grad_norm": 0.1335950493812561, "learning_rate": 1.4592087020026972e-05, "loss": 0.0203, "step": 15260 }, { "grad_norm": 0.11535051465034485, "learning_rate": 1.4533763521386318e-05, "loss": 0.022, "step": 15270 }, { "grad_norm": 0.32268789410591125, "learning_rate": 1.44755369860819e-05, "loss": 0.0185, "step": 15280 }, { "grad_norm": 0.26157623529434204, "learning_rate": 1.441740757330287e-05, "loss": 0.0238, "step": 15290 }, { "grad_norm": 0.3087158799171448, "learning_rate": 1.4359375441972844e-05, "loss": 0.0249, "step": 15300 }, { "grad_norm": 0.15864577889442444, "learning_rate": 1.4301440750749395e-05, "loss": 0.0297, "step": 15310 }, { "grad_norm": 0.15595775842666626, "learning_rate": 1.4243603658023808e-05, "loss": 0.0205, "step": 15320 }, { "grad_norm": 0.3672889471054077, "learning_rate": 1.4185864321920444e-05, "loss": 0.0176, "step": 15330 }, { "grad_norm": 0.11328787356615067, "learning_rate": 1.4128222900296485e-05, "loss": 0.016, "step": 15340 }, { "grad_norm": 0.24173781275749207, "learning_rate": 1.407067955074135e-05, "loss": 0.0254, "step": 15350 }, { "grad_norm": 0.3590124547481537, "learning_rate": 1.4013234430576356e-05, "loss": 0.0155, "step": 15360 }, { "grad_norm": 0.13743779063224792, "learning_rate": 1.3955887696854286e-05, "loss": 0.0251, "step": 15370 }, { "grad_norm": 0.26055940985679626, "learning_rate": 1.38986395063589e-05, "loss": 0.0164, "step": 15380 }, { "grad_norm": 0.41256389021873474, "learning_rate": 1.3841490015604597e-05, "loss": 0.0156, "step": 15390 }, { "grad_norm": 0.13393482565879822, "learning_rate": 1.3784439380835879e-05, "loss": 0.0205, "step": 15400 }, { "grad_norm": 0.20899397134780884, "learning_rate": 1.3727487758026986e-05, "loss": 0.0132, "step": 15410 }, { "grad_norm": 0.19517424702644348, "learning_rate": 1.3670635302881525e-05, "loss": 0.031, "step": 15420 }, { "grad_norm": 0.18729554116725922, "learning_rate": 1.3613882170831888e-05, "loss": 0.0212, "step": 15430 }, { "grad_norm": 0.16335779428482056, "learning_rate": 1.355722851703901e-05, "loss": 0.0253, "step": 15440 }, { "grad_norm": 0.25042128562927246, "learning_rate": 1.3500674496391814e-05, "loss": 0.0197, "step": 15450 }, { "grad_norm": 0.5254186987876892, "learning_rate": 1.3444220263506795e-05, "loss": 0.0169, "step": 15460 }, { "grad_norm": 0.23450221121311188, "learning_rate": 1.3387865972727714e-05, "loss": 0.0182, "step": 15470 }, { "grad_norm": 0.4286744296550751, "learning_rate": 1.3331611778125036e-05, "loss": 0.0317, "step": 15480 }, { "grad_norm": 0.19965983927249908, "learning_rate": 1.3275457833495564e-05, "loss": 0.0194, "step": 15490 }, { "grad_norm": 0.18791216611862183, "learning_rate": 1.3219404292362065e-05, "loss": 0.0187, "step": 15500 }, { "grad_norm": 0.12405511736869812, "learning_rate": 1.3163451307972751e-05, "loss": 0.0191, "step": 15510 }, { "grad_norm": 0.15615221858024597, "learning_rate": 1.3107599033300977e-05, "loss": 0.0142, "step": 15520 }, { "grad_norm": 0.3074841797351837, "learning_rate": 1.305184762104471e-05, "loss": 0.0155, "step": 15530 }, { "grad_norm": 0.3660350739955902, "learning_rate": 1.2996197223626178e-05, "loss": 0.0192, "step": 15540 }, { "grad_norm": 0.22374658286571503, "learning_rate": 1.2940647993191457e-05, "loss": 0.0183, "step": 15550 }, { "grad_norm": 0.2668825089931488, "learning_rate": 1.2885200081610005e-05, "loss": 0.0201, "step": 15560 }, { "grad_norm": 0.35856667160987854, "learning_rate": 1.2829853640474316e-05, "loss": 0.0203, "step": 15570 }, { "grad_norm": 0.17554152011871338, "learning_rate": 1.2774608821099438e-05, "loss": 0.0209, "step": 15580 }, { "grad_norm": 0.12946298718452454, "learning_rate": 1.2719465774522577e-05, "loss": 0.0179, "step": 15590 }, { "grad_norm": 0.09298156946897507, "learning_rate": 1.2664424651502755e-05, "loss": 0.0144, "step": 15600 }, { "grad_norm": 0.20280030369758606, "learning_rate": 1.260948560252026e-05, "loss": 0.027, "step": 15610 }, { "grad_norm": 0.2396259754896164, "learning_rate": 1.2554648777776396e-05, "loss": 0.016, "step": 15620 }, { "grad_norm": 0.2909470498561859, "learning_rate": 1.2499914327192919e-05, "loss": 0.0197, "step": 15630 }, { "grad_norm": 0.1936979740858078, "learning_rate": 1.2445282400411722e-05, "loss": 0.0226, "step": 15640 }, { "grad_norm": 0.41981565952301025, "learning_rate": 1.2390753146794437e-05, "loss": 0.0255, "step": 15650 }, { "grad_norm": 0.40548306703567505, "learning_rate": 1.2336326715421925e-05, "loss": 0.0191, "step": 15660 }, { "grad_norm": 0.18512581288814545, "learning_rate": 1.2282003255094005e-05, "loss": 0.0202, "step": 15670 }, { "grad_norm": 0.14732466638088226, "learning_rate": 1.2227782914328928e-05, "loss": 0.0141, "step": 15680 }, { "grad_norm": 0.32290777564048767, "learning_rate": 1.2173665841363018e-05, "loss": 0.0185, "step": 15690 }, { "grad_norm": 0.40031564235687256, "learning_rate": 1.211965218415032e-05, "loss": 0.0195, "step": 15700 }, { "grad_norm": 0.24879856407642365, "learning_rate": 1.2065742090362082e-05, "loss": 0.0273, "step": 15710 }, { "grad_norm": 0.259458988904953, "learning_rate": 1.2011935707386457e-05, "loss": 0.0097, "step": 15720 }, { "grad_norm": 0.6368922591209412, "learning_rate": 1.1958233182328044e-05, "loss": 0.0286, "step": 15730 }, { "grad_norm": 0.10337379574775696, "learning_rate": 1.1904634662007474e-05, "loss": 0.019, "step": 15740 }, { "grad_norm": 0.15526334941387177, "learning_rate": 1.1851140292961088e-05, "loss": 0.0233, "step": 15750 }, { "grad_norm": 0.18650229275226593, "learning_rate": 1.1797750221440424e-05, "loss": 0.0179, "step": 15760 }, { "grad_norm": 0.2661241292953491, "learning_rate": 1.1744464593411897e-05, "loss": 0.0243, "step": 15770 }, { "grad_norm": 0.12977074086666107, "learning_rate": 1.1691283554556399e-05, "loss": 0.019, "step": 15780 }, { "grad_norm": 0.2002551257610321, "learning_rate": 1.1638207250268834e-05, "loss": 0.0327, "step": 15790 }, { "grad_norm": 0.2548944354057312, "learning_rate": 1.158523582565782e-05, "loss": 0.0212, "step": 15800 }, { "grad_norm": 0.24537190794944763, "learning_rate": 1.1532369425545192e-05, "loss": 0.0251, "step": 15810 }, { "grad_norm": 0.2146523892879486, "learning_rate": 1.1479608194465662e-05, "loss": 0.0229, "step": 15820 }, { "grad_norm": 0.2718653380870819, "learning_rate": 1.1426952276666442e-05, "loss": 0.018, "step": 15830 }, { "grad_norm": 0.18452845513820648, "learning_rate": 1.1374401816106778e-05, "loss": 0.0265, "step": 15840 }, { "grad_norm": 0.558493435382843, "learning_rate": 1.1321956956457646e-05, "loss": 0.0212, "step": 15850 }, { "grad_norm": 0.23108898103237152, "learning_rate": 1.1269617841101277e-05, "loss": 0.0175, "step": 15860 }, { "grad_norm": 0.2189236432313919, "learning_rate": 1.1217384613130804e-05, "loss": 0.0159, "step": 15870 }, { "grad_norm": 0.3419390916824341, "learning_rate": 1.11652574153499e-05, "loss": 0.0184, "step": 15880 }, { "grad_norm": 0.22851309180259705, "learning_rate": 1.1113236390272303e-05, "loss": 0.0214, "step": 15890 }, { "grad_norm": 0.08844264596700668, "learning_rate": 1.106132168012155e-05, "loss": 0.0131, "step": 15900 }, { "grad_norm": 0.22979971766471863, "learning_rate": 1.1009513426830448e-05, "loss": 0.0216, "step": 15910 }, { "grad_norm": 0.3932687044143677, "learning_rate": 1.0957811772040777e-05, "loss": 0.0211, "step": 15920 }, { "grad_norm": 0.21956314146518707, "learning_rate": 1.0906216857102913e-05, "loss": 0.0217, "step": 15930 }, { "grad_norm": 0.15283896028995514, "learning_rate": 1.0854728823075355e-05, "loss": 0.0249, "step": 15940 }, { "grad_norm": 0.13815948367118835, "learning_rate": 1.0803347810724452e-05, "loss": 0.0147, "step": 15950 }, { "grad_norm": 0.2807185649871826, "learning_rate": 1.0752073960523911e-05, "loss": 0.0156, "step": 15960 }, { "grad_norm": 0.492649644613266, "learning_rate": 1.070090741265447e-05, "loss": 0.0154, "step": 15970 }, { "grad_norm": 0.30181267857551575, "learning_rate": 1.0649848307003547e-05, "loss": 0.0139, "step": 15980 }, { "grad_norm": 0.14668402075767517, "learning_rate": 1.0598896783164757e-05, "loss": 0.0261, "step": 15990 }, { "grad_norm": 0.3470149338245392, "learning_rate": 1.0548052980437645e-05, "loss": 0.0175, "step": 16000 }, { "grad_norm": 0.275301992893219, "learning_rate": 1.049731703782722e-05, "loss": 0.023, "step": 16010 }, { "grad_norm": 0.14155562222003937, "learning_rate": 1.0446689094043587e-05, "loss": 0.0205, "step": 16020 }, { "grad_norm": 0.09759993106126785, "learning_rate": 1.039616928750165e-05, "loss": 0.018, "step": 16030 }, { "grad_norm": 0.17028358578681946, "learning_rate": 1.0345757756320612e-05, "loss": 0.0168, "step": 16040 }, { "grad_norm": 0.2885197103023529, "learning_rate": 1.0295454638323666e-05, "loss": 0.0209, "step": 16050 }, { "grad_norm": 0.414573073387146, "learning_rate": 1.0245260071037632e-05, "loss": 0.0222, "step": 16060 }, { "grad_norm": 0.34932151436805725, "learning_rate": 1.0195174191692518e-05, "loss": 0.0169, "step": 16070 }, { "grad_norm": 0.3162248134613037, "learning_rate": 1.014519713722124e-05, "loss": 0.0218, "step": 16080 }, { "grad_norm": 0.2436322569847107, "learning_rate": 1.0095329044259132e-05, "loss": 0.0183, "step": 16090 }, { "grad_norm": 0.1829928457736969, "learning_rate": 1.004557004914365e-05, "loss": 0.0187, "step": 16100 }, { "grad_norm": 0.21037498116493225, "learning_rate": 9.995920287914007e-06, "loss": 0.0182, "step": 16110 }, { "grad_norm": 0.33513617515563965, "learning_rate": 9.946379896310737e-06, "loss": 0.0225, "step": 16120 }, { "grad_norm": 0.47261056303977966, "learning_rate": 9.896949009775396e-06, "loss": 0.024, "step": 16130 }, { "grad_norm": 0.27161362767219543, "learning_rate": 9.847627763450134e-06, "loss": 0.0185, "step": 16140 }, { "grad_norm": 0.08532784879207611, "learning_rate": 9.798416292177337e-06, "loss": 0.0208, "step": 16150 }, { "grad_norm": 0.1765655130147934, "learning_rate": 9.74931473049932e-06, "loss": 0.0263, "step": 16160 }, { "grad_norm": 0.30696067214012146, "learning_rate": 9.700323212657847e-06, "loss": 0.0162, "step": 16170 }, { "grad_norm": 0.14078685641288757, "learning_rate": 9.65144187259388e-06, "loss": 0.0113, "step": 16180 }, { "grad_norm": 0.5275670886039734, "learning_rate": 9.602670843947132e-06, "loss": 0.0222, "step": 16190 }, { "grad_norm": 0.12976324558258057, "learning_rate": 9.554010260055713e-06, "loss": 0.0212, "step": 16200 }, { "grad_norm": 0.16918262839317322, "learning_rate": 9.505460253955834e-06, "loss": 0.026, "step": 16210 }, { "grad_norm": 0.13150447607040405, "learning_rate": 9.457020958381324e-06, "loss": 0.0131, "step": 16220 }, { "grad_norm": 0.06802402436733246, "learning_rate": 9.408692505763395e-06, "loss": 0.0208, "step": 16230 }, { "grad_norm": 0.31744250655174255, "learning_rate": 9.360475028230181e-06, "loss": 0.0204, "step": 16240 }, { "grad_norm": 0.14333786070346832, "learning_rate": 9.312368657606412e-06, "loss": 0.0254, "step": 16250 }, { "grad_norm": 0.34759268164634705, "learning_rate": 9.264373525413096e-06, "loss": 0.019, "step": 16260 }, { "grad_norm": 0.12161330878734589, "learning_rate": 9.216489762867058e-06, "loss": 0.0227, "step": 16270 }, { "grad_norm": 0.20595809817314148, "learning_rate": 9.168717500880708e-06, "loss": 0.0217, "step": 16280 }, { "grad_norm": 0.06790205836296082, "learning_rate": 9.121056870061574e-06, "loss": 0.0171, "step": 16290 }, { "grad_norm": 0.17432934045791626, "learning_rate": 9.073508000711983e-06, "loss": 0.0168, "step": 16300 }, { "grad_norm": 0.3055286705493927, "learning_rate": 9.026071022828758e-06, "loss": 0.0221, "step": 16310 }, { "grad_norm": 0.16395656764507294, "learning_rate": 8.978746066102771e-06, "loss": 0.0198, "step": 16320 }, { "grad_norm": 0.1211945042014122, "learning_rate": 8.931533259918634e-06, "loss": 0.0174, "step": 16330 }, { "grad_norm": 0.14405213296413422, "learning_rate": 8.884432733354382e-06, "loss": 0.0132, "step": 16340 }, { "grad_norm": 0.14549356698989868, "learning_rate": 8.837444615181029e-06, "loss": 0.0165, "step": 16350 }, { "grad_norm": 0.2788192629814148, "learning_rate": 8.790569033862323e-06, "loss": 0.025, "step": 16360 }, { "grad_norm": 0.19466930627822876, "learning_rate": 8.7438061175543e-06, "loss": 0.0313, "step": 16370 }, { "grad_norm": 0.1067361831665039, "learning_rate": 8.697155994104978e-06, "loss": 0.0184, "step": 16380 }, { "grad_norm": 0.14124664664268494, "learning_rate": 8.650618791054033e-06, "loss": 0.0267, "step": 16390 }, { "grad_norm": 0.19008152186870575, "learning_rate": 8.604194635632373e-06, "loss": 0.0254, "step": 16400 }, { "grad_norm": 0.3366881012916565, "learning_rate": 8.557883654761906e-06, "loss": 0.0128, "step": 16410 }, { "grad_norm": 0.18231715261936188, "learning_rate": 8.511685975055061e-06, "loss": 0.0141, "step": 16420 }, { "grad_norm": 0.3234505355358124, "learning_rate": 8.46560172281452e-06, "loss": 0.0208, "step": 16430 }, { "grad_norm": 0.11903863400220871, "learning_rate": 8.419631024032893e-06, "loss": 0.0174, "step": 16440 }, { "grad_norm": 0.15012577176094055, "learning_rate": 8.373774004392293e-06, "loss": 0.0117, "step": 16450 }, { "grad_norm": 0.30577418208122253, "learning_rate": 8.32803078926409e-06, "loss": 0.0188, "step": 16460 }, { "grad_norm": 0.18174400925636292, "learning_rate": 8.282401503708454e-06, "loss": 0.0143, "step": 16470 }, { "grad_norm": 0.2725290060043335, "learning_rate": 8.23688627247412e-06, "loss": 0.0221, "step": 16480 }, { "grad_norm": 0.16705700755119324, "learning_rate": 8.191485219998007e-06, "loss": 0.0233, "step": 16490 }, { "grad_norm": 0.3855743706226349, "learning_rate": 8.146198470404843e-06, "loss": 0.015, "step": 16500 }, { "grad_norm": 0.256012886762619, "learning_rate": 8.101026147506897e-06, "loss": 0.0151, "step": 16510 }, { "grad_norm": 0.0904875248670578, "learning_rate": 8.05596837480353e-06, "loss": 0.0252, "step": 16520 }, { "grad_norm": 0.2854001224040985, "learning_rate": 8.011025275480998e-06, "loss": 0.0217, "step": 16530 }, { "grad_norm": 0.3644357919692993, "learning_rate": 7.966196972412027e-06, "loss": 0.0166, "step": 16540 }, { "grad_norm": 0.2792627215385437, "learning_rate": 7.92148358815547e-06, "loss": 0.0176, "step": 16550 }, { "grad_norm": 0.3277478814125061, "learning_rate": 7.87688524495604e-06, "loss": 0.0248, "step": 16560 }, { "grad_norm": 0.14974917471408844, "learning_rate": 7.83240206474386e-06, "loss": 0.0166, "step": 16570 }, { "grad_norm": 0.12707270681858063, "learning_rate": 7.788034169134272e-06, "loss": 0.0153, "step": 16580 }, { "grad_norm": 0.1138826459646225, "learning_rate": 7.743781679427414e-06, "loss": 0.016, "step": 16590 }, { "grad_norm": 0.12539009749889374, "learning_rate": 7.699644716607895e-06, "loss": 0.0118, "step": 16600 }, { "grad_norm": 0.1484643518924713, "learning_rate": 7.655623401344486e-06, "loss": 0.0134, "step": 16610 }, { "grad_norm": 0.2696562707424164, "learning_rate": 7.611717853989775e-06, "loss": 0.0158, "step": 16620 }, { "grad_norm": 0.19487886130809784, "learning_rate": 7.567928194579854e-06, "loss": 0.0136, "step": 16630 }, { "grad_norm": 0.41545915603637695, "learning_rate": 7.524254542833997e-06, "loss": 0.0211, "step": 16640 }, { "grad_norm": 0.19425921142101288, "learning_rate": 7.480697018154286e-06, "loss": 0.0238, "step": 16650 }, { "grad_norm": 0.14019429683685303, "learning_rate": 7.437255739625332e-06, "loss": 0.0129, "step": 16660 }, { "grad_norm": 0.20017379522323608, "learning_rate": 7.393930826013923e-06, "loss": 0.0322, "step": 16670 }, { "grad_norm": 0.27543213963508606, "learning_rate": 7.350722395768722e-06, "loss": 0.0209, "step": 16680 }, { "grad_norm": 0.21665240824222565, "learning_rate": 7.307630567019963e-06, "loss": 0.022, "step": 16690 }, { "grad_norm": 0.17671620845794678, "learning_rate": 7.264655457579e-06, "loss": 0.0164, "step": 16700 }, { "grad_norm": 0.20855896174907684, "learning_rate": 7.221797184938184e-06, "loss": 0.019, "step": 16710 }, { "grad_norm": 0.2924117147922516, "learning_rate": 7.179055866270373e-06, "loss": 0.025, "step": 16720 }, { "grad_norm": 0.26090070605278015, "learning_rate": 7.136431618428707e-06, "loss": 0.0233, "step": 16730 }, { "grad_norm": 0.17672841250896454, "learning_rate": 7.09392455794628e-06, "loss": 0.02, "step": 16740 }, { "grad_norm": 0.14410196244716644, "learning_rate": 7.051534801035725e-06, "loss": 0.0201, "step": 16750 }, { "grad_norm": 0.30365467071533203, "learning_rate": 7.00926246358905e-06, "loss": 0.0168, "step": 16760 }, { "grad_norm": 0.16273841261863708, "learning_rate": 6.967107661177191e-06, "loss": 0.0167, "step": 16770 }, { "grad_norm": 0.28842324018478394, "learning_rate": 6.925070509049786e-06, "loss": 0.0126, "step": 16780 }, { "grad_norm": 0.13757379353046417, "learning_rate": 6.883151122134812e-06, "loss": 0.012, "step": 16790 }, { "grad_norm": 0.2168184071779251, "learning_rate": 6.8413496150382394e-06, "loss": 0.0195, "step": 16800 }, { "grad_norm": 0.2700727581977844, "learning_rate": 6.7996661020438165e-06, "loss": 0.0161, "step": 16810 }, { "grad_norm": 0.18629080057144165, "learning_rate": 6.758100697112662e-06, "loss": 0.0117, "step": 16820 }, { "grad_norm": 0.2896629869937897, "learning_rate": 6.716653513883026e-06, "loss": 0.016, "step": 16830 }, { "grad_norm": 0.20963476598262787, "learning_rate": 6.675324665669913e-06, "loss": 0.0163, "step": 16840 }, { "grad_norm": 0.29005515575408936, "learning_rate": 6.634114265464803e-06, "loss": 0.0177, "step": 16850 }, { "grad_norm": 0.1386127471923828, "learning_rate": 6.59302242593538e-06, "loss": 0.0183, "step": 16860 }, { "grad_norm": 0.1391477882862091, "learning_rate": 6.552049259425141e-06, "loss": 0.0131, "step": 16870 }, { "grad_norm": 0.22799283266067505, "learning_rate": 6.511194877953181e-06, "loss": 0.0196, "step": 16880 }, { "grad_norm": 0.31572994589805603, "learning_rate": 6.470459393213813e-06, "loss": 0.0199, "step": 16890 }, { "grad_norm": 0.23816099762916565, "learning_rate": 6.429842916576279e-06, "loss": 0.0132, "step": 16900 }, { "grad_norm": 0.1799507737159729, "learning_rate": 6.389345559084503e-06, "loss": 0.0148, "step": 16910 }, { "grad_norm": 0.1961701661348343, "learning_rate": 6.348967431456682e-06, "loss": 0.0183, "step": 16920 }, { "grad_norm": 0.35844114422798157, "learning_rate": 6.30870864408511e-06, "loss": 0.0212, "step": 16930 }, { "grad_norm": 0.13372072577476501, "learning_rate": 6.268569307035754e-06, "loss": 0.0201, "step": 16940 }, { "grad_norm": 0.20245884358882904, "learning_rate": 6.228549530048022e-06, "loss": 0.0313, "step": 16950 }, { "grad_norm": 0.3614831566810608, "learning_rate": 6.1886494225344814e-06, "loss": 0.0215, "step": 16960 }, { "grad_norm": 0.13196921348571777, "learning_rate": 6.148869093580479e-06, "loss": 0.0171, "step": 16970 }, { "grad_norm": 0.1699007898569107, "learning_rate": 6.109208651943921e-06, "loss": 0.0227, "step": 16980 }, { "grad_norm": 0.20307552814483643, "learning_rate": 6.069668206054946e-06, "loss": 0.0142, "step": 16990 }, { "grad_norm": 0.5556154847145081, "learning_rate": 6.0302478640156145e-06, "loss": 0.0275, "step": 17000 }, { "grad_norm": 0.30989179015159607, "learning_rate": 5.990947733599644e-06, "loss": 0.0261, "step": 17010 }, { "grad_norm": 0.2558532655239105, "learning_rate": 5.951767922252105e-06, "loss": 0.0186, "step": 17020 }, { "grad_norm": 0.13263508677482605, "learning_rate": 5.912708537089068e-06, "loss": 0.0098, "step": 17030 }, { "grad_norm": 0.16918788850307465, "learning_rate": 5.873769684897434e-06, "loss": 0.0167, "step": 17040 }, { "grad_norm": 0.3015671372413635, "learning_rate": 5.834951472134514e-06, "loss": 0.0246, "step": 17050 }, { "grad_norm": 0.45994654297828674, "learning_rate": 5.796254004927832e-06, "loss": 0.0202, "step": 17060 }, { "grad_norm": 0.06840777397155762, "learning_rate": 5.757677389074806e-06, "loss": 0.0133, "step": 17070 }, { "grad_norm": 0.24574226140975952, "learning_rate": 5.719221730042385e-06, "loss": 0.0118, "step": 17080 }, { "grad_norm": 0.24014350771903992, "learning_rate": 5.680887132966911e-06, "loss": 0.0178, "step": 17090 }, { "grad_norm": 0.1888381689786911, "learning_rate": 5.642673702653683e-06, "loss": 0.0141, "step": 17100 }, { "grad_norm": 0.318472295999527, "learning_rate": 5.604581543576781e-06, "loss": 0.02, "step": 17110 }, { "grad_norm": 0.18264414370059967, "learning_rate": 5.566610759878704e-06, "loss": 0.021, "step": 17120 }, { "grad_norm": 0.21229355037212372, "learning_rate": 5.528761455370119e-06, "loss": 0.0135, "step": 17130 }, { "grad_norm": 0.2837013602256775, "learning_rate": 5.491033733529594e-06, "loss": 0.0124, "step": 17140 }, { "grad_norm": 0.16953051090240479, "learning_rate": 5.453427697503255e-06, "loss": 0.0221, "step": 17150 }, { "grad_norm": 0.17823177576065063, "learning_rate": 5.415943450104599e-06, "loss": 0.0221, "step": 17160 }, { "grad_norm": 0.3249962031841278, "learning_rate": 5.378581093814111e-06, "loss": 0.0216, "step": 17170 }, { "grad_norm": 0.8187869787216187, "learning_rate": 5.3413407307790375e-06, "loss": 0.0127, "step": 17180 }, { "grad_norm": 0.3042313754558563, "learning_rate": 5.30422246281313e-06, "loss": 0.0153, "step": 17190 }, { "grad_norm": 0.16173231601715088, "learning_rate": 5.267226391396296e-06, "loss": 0.0133, "step": 17200 }, { "grad_norm": 0.22072267532348633, "learning_rate": 5.2303526176744e-06, "loss": 0.0215, "step": 17210 }, { "grad_norm": 0.11511935293674469, "learning_rate": 5.193601242458929e-06, "loss": 0.0195, "step": 17220 }, { "grad_norm": 0.27928370237350464, "learning_rate": 5.156972366226714e-06, "loss": 0.0131, "step": 17230 }, { "grad_norm": 0.2095157653093338, "learning_rate": 5.120466089119735e-06, "loss": 0.0172, "step": 17240 }, { "grad_norm": 0.16289864480495453, "learning_rate": 5.084082510944749e-06, "loss": 0.0361, "step": 17250 }, { "grad_norm": 0.3902093172073364, "learning_rate": 5.047821731173058e-06, "loss": 0.0146, "step": 17260 }, { "grad_norm": 0.3603057563304901, "learning_rate": 5.011683848940274e-06, "loss": 0.0245, "step": 17270 }, { "grad_norm": 0.17812711000442505, "learning_rate": 4.975668963045954e-06, "loss": 0.0151, "step": 17280 }, { "grad_norm": 0.15019574761390686, "learning_rate": 4.9397771719534525e-06, "loss": 0.0152, "step": 17290 }, { "grad_norm": 0.21430954337120056, "learning_rate": 4.904008573789548e-06, "loss": 0.0155, "step": 17300 }, { "grad_norm": 0.21786090731620789, "learning_rate": 4.8683632663442005e-06, "loss": 0.0157, "step": 17310 }, { "grad_norm": 0.08686906844377518, "learning_rate": 4.832841347070343e-06, "loss": 0.0119, "step": 17320 }, { "grad_norm": 0.10057278722524643, "learning_rate": 4.797442913083539e-06, "loss": 0.0102, "step": 17330 }, { "grad_norm": 0.14532889425754547, "learning_rate": 4.7621680611617596e-06, "loss": 0.0188, "step": 17340 }, { "grad_norm": 0.19129155576229095, "learning_rate": 4.727016887745095e-06, "loss": 0.0116, "step": 17350 }, { "grad_norm": 0.5408861041069031, "learning_rate": 4.691989488935511e-06, "loss": 0.0177, "step": 17360 }, { "grad_norm": 0.23207971453666687, "learning_rate": 4.657085960496588e-06, "loss": 0.0171, "step": 17370 }, { "grad_norm": 0.3333640694618225, "learning_rate": 4.6223063978532265e-06, "loss": 0.0156, "step": 17380 }, { "grad_norm": 0.15695223212242126, "learning_rate": 4.587650896091439e-06, "loss": 0.016, "step": 17390 }, { "grad_norm": 0.21006596088409424, "learning_rate": 4.553119549958035e-06, "loss": 0.0194, "step": 17400 }, { "grad_norm": 0.17558525502681732, "learning_rate": 4.518712453860385e-06, "loss": 0.0114, "step": 17410 }, { "grad_norm": 0.355133593082428, "learning_rate": 4.484429701866205e-06, "loss": 0.0193, "step": 17420 }, { "grad_norm": 0.1898612082004547, "learning_rate": 4.4502713877031975e-06, "loss": 0.0137, "step": 17430 }, { "grad_norm": 0.22224687039852142, "learning_rate": 4.416237604758911e-06, "loss": 0.0215, "step": 17440 }, { "grad_norm": 0.21468329429626465, "learning_rate": 4.3823284460804025e-06, "loss": 0.0154, "step": 17450 }, { "grad_norm": 0.09784099459648132, "learning_rate": 4.348544004374011e-06, "loss": 0.0125, "step": 17460 }, { "grad_norm": 0.14415784180164337, "learning_rate": 4.314884372005123e-06, "loss": 0.0092, "step": 17470 }, { "grad_norm": 0.15692779421806335, "learning_rate": 4.281349640997867e-06, "loss": 0.0227, "step": 17480 }, { "grad_norm": 0.13794566690921783, "learning_rate": 4.247939903034942e-06, "loss": 0.0111, "step": 17490 }, { "grad_norm": 0.16269980370998383, "learning_rate": 4.214655249457284e-06, "loss": 0.0158, "step": 17500 }, { "grad_norm": 0.49687787890434265, "learning_rate": 4.181495771263855e-06, "loss": 0.0217, "step": 17510 }, { "grad_norm": 0.08058241754770279, "learning_rate": 4.148461559111427e-06, "loss": 0.0186, "step": 17520 }, { "grad_norm": 0.15929141640663147, "learning_rate": 4.115552703314252e-06, "loss": 0.0199, "step": 17530 }, { "grad_norm": 0.13761360943317413, "learning_rate": 4.082769293843886e-06, "loss": 0.0217, "step": 17540 }, { "grad_norm": 0.14527955651283264, "learning_rate": 4.050111420328939e-06, "loss": 0.0162, "step": 17550 }, { "grad_norm": 0.3188728392124176, "learning_rate": 4.017579172054764e-06, "loss": 0.0325, "step": 17560 }, { "grad_norm": 0.15149231255054474, "learning_rate": 3.985172637963308e-06, "loss": 0.0184, "step": 17570 }, { "grad_norm": 0.2080678939819336, "learning_rate": 3.952891906652784e-06, "loss": 0.0178, "step": 17580 }, { "grad_norm": 0.23599326610565186, "learning_rate": 3.920737066377478e-06, "loss": 0.0191, "step": 17590 }, { "grad_norm": 0.200724795460701, "learning_rate": 3.888708205047509e-06, "loss": 0.0258, "step": 17600 }, { "grad_norm": 0.3828006386756897, "learning_rate": 3.856805410228542e-06, "loss": 0.0216, "step": 17610 }, { "grad_norm": 0.18195831775665283, "learning_rate": 3.82502876914162e-06, "loss": 0.0242, "step": 17620 }, { "grad_norm": 0.16787761449813843, "learning_rate": 3.7933783686628586e-06, "loss": 0.0116, "step": 17630 }, { "grad_norm": 0.26111480593681335, "learning_rate": 3.7618542953232306e-06, "loss": 0.0217, "step": 17640 }, { "grad_norm": 0.08528207987546921, "learning_rate": 3.7304566353083658e-06, "loss": 0.0187, "step": 17650 }, { "grad_norm": 0.45704835653305054, "learning_rate": 3.6991854744582555e-06, "loss": 0.034, "step": 17660 }, { "grad_norm": 0.24210789799690247, "learning_rate": 3.6680408982670777e-06, "loss": 0.0168, "step": 17670 }, { "grad_norm": 0.32769036293029785, "learning_rate": 3.637022991882899e-06, "loss": 0.0181, "step": 17680 }, { "grad_norm": 0.15214061737060547, "learning_rate": 3.606131840107485e-06, "loss": 0.01, "step": 17690 }, { "grad_norm": 0.20557653903961182, "learning_rate": 3.575367527396084e-06, "loss": 0.0242, "step": 17700 }, { "grad_norm": 0.33930450677871704, "learning_rate": 3.5447301378571386e-06, "loss": 0.0273, "step": 17710 }, { "grad_norm": 0.20265765488147736, "learning_rate": 3.514219755252113e-06, "loss": 0.0121, "step": 17720 }, { "grad_norm": 0.28554826974868774, "learning_rate": 3.4838364629952213e-06, "loss": 0.0194, "step": 17730 }, { "grad_norm": 0.34949299693107605, "learning_rate": 3.4535803441532123e-06, "loss": 0.0175, "step": 17740 }, { "grad_norm": 0.31058773398399353, "learning_rate": 3.4234514814451836e-06, "loss": 0.0271, "step": 17750 }, { "grad_norm": 0.22975371778011322, "learning_rate": 3.393449957242273e-06, "loss": 0.0188, "step": 17760 }, { "grad_norm": 0.1925979107618332, "learning_rate": 3.363575853567524e-06, "loss": 0.0202, "step": 17770 }, { "grad_norm": 0.34745049476623535, "learning_rate": 3.3338292520955826e-06, "loss": 0.0257, "step": 17780 }, { "grad_norm": 0.06988287717103958, "learning_rate": 3.304210234152516e-06, "loss": 0.0111, "step": 17790 }, { "grad_norm": 0.30880123376846313, "learning_rate": 3.2747188807155993e-06, "loss": 0.0112, "step": 17800 }, { "grad_norm": 0.2922608554363251, "learning_rate": 3.2453552724130643e-06, "loss": 0.0183, "step": 17810 }, { "grad_norm": 0.38370999693870544, "learning_rate": 3.216119489523889e-06, "loss": 0.013, "step": 17820 }, { "grad_norm": 0.2571949064731598, "learning_rate": 3.1870116119775917e-06, "loss": 0.0166, "step": 17830 }, { "grad_norm": 0.26999253034591675, "learning_rate": 3.158031719353999e-06, "loss": 0.0199, "step": 17840 }, { "grad_norm": 0.1487848460674286, "learning_rate": 3.1291798908830273e-06, "loss": 0.0143, "step": 17850 }, { "grad_norm": 0.18769730627536774, "learning_rate": 3.1004562054444853e-06, "loss": 0.015, "step": 17860 }, { "grad_norm": 0.327690064907074, "learning_rate": 3.071860741567806e-06, "loss": 0.0124, "step": 17870 }, { "grad_norm": 0.3256794214248657, "learning_rate": 3.04339357743193e-06, "loss": 0.0125, "step": 17880 }, { "grad_norm": 0.1420399397611618, "learning_rate": 3.0150547908649628e-06, "loss": 0.0219, "step": 17890 }, { "grad_norm": 0.2030663937330246, "learning_rate": 2.9868444593440957e-06, "loss": 0.0107, "step": 17900 }, { "grad_norm": 0.4686635136604309, "learning_rate": 2.9587626599952846e-06, "loss": 0.0152, "step": 17910 }, { "grad_norm": 0.09315881133079529, "learning_rate": 2.930809469593082e-06, "loss": 0.0159, "step": 17920 }, { "grad_norm": 0.2677760422229767, "learning_rate": 2.9029849645604733e-06, "loss": 0.0189, "step": 17930 }, { "grad_norm": 0.17986221611499786, "learning_rate": 2.8752892209685632e-06, "loss": 0.0147, "step": 17940 }, { "grad_norm": 0.05794787034392357, "learning_rate": 2.847722314536483e-06, "loss": 0.0217, "step": 17950 }, { "grad_norm": 0.2953844666481018, "learning_rate": 2.820284320631078e-06, "loss": 0.0203, "step": 17960 }, { "grad_norm": 0.1307225525379181, "learning_rate": 2.792975314266788e-06, "loss": 0.0127, "step": 17970 }, { "grad_norm": 0.2128330022096634, "learning_rate": 2.7657953701054007e-06, "loss": 0.0218, "step": 17980 }, { "grad_norm": 0.19662301242351532, "learning_rate": 2.7387445624558306e-06, "loss": 0.0136, "step": 17990 }, { "grad_norm": 0.16931232810020447, "learning_rate": 2.7118229652739747e-06, "loss": 0.0158, "step": 18000 }, { "grad_norm": 0.32883375883102417, "learning_rate": 2.6850306521624236e-06, "loss": 0.0309, "step": 18010 }, { "grad_norm": 0.2620268166065216, "learning_rate": 2.6583676963703507e-06, "loss": 0.0232, "step": 18020 }, { "grad_norm": 0.1539437174797058, "learning_rate": 2.631834170793268e-06, "loss": 0.0162, "step": 18030 }, { "grad_norm": 0.41769349575042725, "learning_rate": 2.6054301479728036e-06, "loss": 0.0184, "step": 18040 }, { "grad_norm": 0.48504552245140076, "learning_rate": 2.579155700096575e-06, "loss": 0.0219, "step": 18050 }, { "grad_norm": 0.5269713997840881, "learning_rate": 2.5530108989978873e-06, "loss": 0.0224, "step": 18060 }, { "grad_norm": 0.27721697092056274, "learning_rate": 2.5269958161556416e-06, "loss": 0.0169, "step": 18070 }, { "grad_norm": 0.27810049057006836, "learning_rate": 2.5011105226940888e-06, "loss": 0.013, "step": 18080 }, { "grad_norm": 0.14821752905845642, "learning_rate": 2.4753550893826248e-06, "loss": 0.015, "step": 18090 }, { "grad_norm": 0.08294545859098434, "learning_rate": 2.4497295866356296e-06, "loss": 0.0222, "step": 18100 }, { "grad_norm": 0.3570573925971985, "learning_rate": 2.424234084512228e-06, "loss": 0.0198, "step": 18110 }, { "grad_norm": 0.2957823872566223, "learning_rate": 2.3988686527161687e-06, "loss": 0.0174, "step": 18120 }, { "grad_norm": 0.38920310139656067, "learning_rate": 2.373633360595573e-06, "loss": 0.0173, "step": 18130 }, { "grad_norm": 0.070996955037117, "learning_rate": 2.3485282771427585e-06, "loss": 0.0114, "step": 18140 }, { "grad_norm": 0.6089935302734375, "learning_rate": 2.3235534709940665e-06, "loss": 0.0137, "step": 18150 }, { "grad_norm": 0.20129522681236267, "learning_rate": 2.2987090104296617e-06, "loss": 0.0178, "step": 18160 }, { "grad_norm": 0.26319703459739685, "learning_rate": 2.273994963373355e-06, "loss": 0.021, "step": 18170 }, { "grad_norm": 0.4135395884513855, "learning_rate": 2.249411397392409e-06, "loss": 0.0152, "step": 18180 }, { "grad_norm": 0.09135359525680542, "learning_rate": 2.2249583796973506e-06, "loss": 0.0194, "step": 18190 }, { "grad_norm": 0.12823326885700226, "learning_rate": 2.200635977141796e-06, "loss": 0.0135, "step": 18200 }, { "grad_norm": 0.16674955189228058, "learning_rate": 2.17644425622226e-06, "loss": 0.0234, "step": 18210 }, { "grad_norm": 0.11586598306894302, "learning_rate": 2.152383283077991e-06, "loss": 0.0115, "step": 18220 }, { "grad_norm": 0.06953791528940201, "learning_rate": 2.128453123490781e-06, "loss": 0.0177, "step": 18230 }, { "grad_norm": 0.3114459216594696, "learning_rate": 2.1046538428847462e-06, "loss": 0.0158, "step": 18240 }, { "grad_norm": 0.28075313568115234, "learning_rate": 2.0809855063262273e-06, "loss": 0.0141, "step": 18250 }, { "grad_norm": 0.22420334815979004, "learning_rate": 2.057448178523558e-06, "loss": 0.0124, "step": 18260 }, { "grad_norm": 0.37695497274398804, "learning_rate": 2.034041923826885e-06, "loss": 0.0149, "step": 18270 }, { "grad_norm": 0.17755354940891266, "learning_rate": 2.0107668062280204e-06, "loss": 0.014, "step": 18280 }, { "grad_norm": 0.30442553758621216, "learning_rate": 1.9876228893602357e-06, "loss": 0.0206, "step": 18290 }, { "grad_norm": 0.23252388834953308, "learning_rate": 1.9646102364981266e-06, "loss": 0.0153, "step": 18300 }, { "grad_norm": 0.08676865696907043, "learning_rate": 1.9417289105574053e-06, "loss": 0.0183, "step": 18310 }, { "grad_norm": 0.3853219747543335, "learning_rate": 1.9189789740947427e-06, "loss": 0.0232, "step": 18320 }, { "grad_norm": 0.10027401894330978, "learning_rate": 1.896360489307597e-06, "loss": 0.0124, "step": 18330 }, { "grad_norm": 0.36713874340057373, "learning_rate": 1.8738735180340362e-06, "loss": 0.0187, "step": 18340 }, { "grad_norm": 0.19531823694705963, "learning_rate": 1.8515181217525824e-06, "loss": 0.0105, "step": 18350 }, { "grad_norm": 0.2512531578540802, "learning_rate": 1.8292943615820457e-06, "loss": 0.0189, "step": 18360 }, { "grad_norm": 0.3074987232685089, "learning_rate": 1.8072022982813296e-06, "loss": 0.017, "step": 18370 }, { "grad_norm": 0.23885013163089752, "learning_rate": 1.7852419922492925e-06, "loss": 0.0205, "step": 18380 }, { "grad_norm": 0.07534903287887573, "learning_rate": 1.763413503524569e-06, "loss": 0.0138, "step": 18390 }, { "grad_norm": 0.3960243761539459, "learning_rate": 1.7417168917854165e-06, "loss": 0.0284, "step": 18400 }, { "grad_norm": 0.3367649018764496, "learning_rate": 1.720152216349552e-06, "loss": 0.0162, "step": 18410 }, { "grad_norm": 0.19240038096904755, "learning_rate": 1.6987195361739595e-06, "loss": 0.0217, "step": 18420 }, { "grad_norm": 0.07945838570594788, "learning_rate": 1.6774189098547832e-06, "loss": 0.0156, "step": 18430 }, { "grad_norm": 0.29379066824913025, "learning_rate": 1.6562503956271069e-06, "loss": 0.0096, "step": 18440 }, { "grad_norm": 0.4338149130344391, "learning_rate": 1.6352140513648417e-06, "loss": 0.0212, "step": 18450 }, { "grad_norm": 0.43205925822257996, "learning_rate": 1.6143099345805712e-06, "loss": 0.0178, "step": 18460 }, { "grad_norm": 0.4978521466255188, "learning_rate": 1.5935381024253293e-06, "loss": 0.0172, "step": 18470 }, { "grad_norm": 0.11142927408218384, "learning_rate": 1.572898611688517e-06, "loss": 0.0146, "step": 18480 }, { "grad_norm": 0.6844159364700317, "learning_rate": 1.5523915187977133e-06, "loss": 0.0168, "step": 18490 }, { "grad_norm": 0.1689697802066803, "learning_rate": 1.532016879818532e-06, "loss": 0.025, "step": 18500 }, { "grad_norm": 0.19397972524166107, "learning_rate": 1.51177475045447e-06, "loss": 0.0191, "step": 18510 }, { "grad_norm": 0.2852877080440521, "learning_rate": 1.4916651860467035e-06, "loss": 0.0133, "step": 18520 }, { "grad_norm": 0.27902641892433167, "learning_rate": 1.471688241574043e-06, "loss": 0.0238, "step": 18530 }, { "grad_norm": 0.18983925879001617, "learning_rate": 1.451843971652672e-06, "loss": 0.0119, "step": 18540 }, { "grad_norm": 0.28655335307121277, "learning_rate": 1.432132430536076e-06, "loss": 0.0158, "step": 18550 }, { "grad_norm": 0.2359672337770462, "learning_rate": 1.412553672114869e-06, "loss": 0.0236, "step": 18560 }, { "grad_norm": 0.3403007984161377, "learning_rate": 1.3931077499166056e-06, "loss": 0.0224, "step": 18570 }, { "grad_norm": 0.4586842954158783, "learning_rate": 1.3737947171057085e-06, "loss": 0.0137, "step": 18580 }, { "grad_norm": 0.09274456650018692, "learning_rate": 1.3546146264832582e-06, "loss": 0.0115, "step": 18590 }, { "grad_norm": 0.452818363904953, "learning_rate": 1.3355675304869086e-06, "loss": 0.0292, "step": 18600 }, { "grad_norm": 0.27036717534065247, "learning_rate": 1.3166534811906827e-06, "loss": 0.0178, "step": 18610 }, { "grad_norm": 0.41751429438591003, "learning_rate": 1.2978725303048666e-06, "loss": 0.0253, "step": 18620 }, { "grad_norm": 0.43862372636795044, "learning_rate": 1.2792247291758762e-06, "loss": 0.0325, "step": 18630 }, { "grad_norm": 0.3639572262763977, "learning_rate": 1.2607101287860635e-06, "loss": 0.016, "step": 18640 }, { "grad_norm": 0.23536613583564758, "learning_rate": 1.2423287797536654e-06, "loss": 0.0171, "step": 18650 }, { "grad_norm": 0.15410111844539642, "learning_rate": 1.2240807323325776e-06, "loss": 0.0204, "step": 18660 }, { "grad_norm": 0.3498094379901886, "learning_rate": 1.205966036412254e-06, "loss": 0.0217, "step": 18670 }, { "grad_norm": 0.12219243496656418, "learning_rate": 1.1879847415175949e-06, "loss": 0.0147, "step": 18680 }, { "grad_norm": 0.2015448361635208, "learning_rate": 1.1701368968087712e-06, "loss": 0.0148, "step": 18690 }, { "grad_norm": 0.1641262024641037, "learning_rate": 1.1524225510811116e-06, "loss": 0.014, "step": 18700 }, { "grad_norm": 0.0829595997929573, "learning_rate": 1.1348417527649535e-06, "loss": 0.018, "step": 18710 }, { "grad_norm": 0.17638997733592987, "learning_rate": 1.1173945499255268e-06, "loss": 0.0151, "step": 18720 }, { "grad_norm": 0.24638019502162933, "learning_rate": 1.1000809902628307e-06, "loss": 0.0345, "step": 18730 }, { "grad_norm": 0.24256514012813568, "learning_rate": 1.082901121111468e-06, "loss": 0.0179, "step": 18740 }, { "grad_norm": 0.17310838401317596, "learning_rate": 1.0658549894405456e-06, "loss": 0.0155, "step": 18750 }, { "grad_norm": 0.22420494258403778, "learning_rate": 1.0489426418535342e-06, "loss": 0.0141, "step": 18760 }, { "grad_norm": 0.05709094926714897, "learning_rate": 1.0321641245881474e-06, "loss": 0.0246, "step": 18770 }, { "grad_norm": 0.33535465598106384, "learning_rate": 1.015519483516214e-06, "loss": 0.0194, "step": 18780 }, { "grad_norm": 0.4891705811023712, "learning_rate": 9.990087641435443e-07, "loss": 0.018, "step": 18790 }, { "grad_norm": 0.22315281629562378, "learning_rate": 9.826320116098132e-07, "loss": 0.0169, "step": 18800 }, { "grad_norm": 0.28896740078926086, "learning_rate": 9.663892706884447e-07, "loss": 0.0152, "step": 18810 }, { "grad_norm": 0.3949716091156006, "learning_rate": 9.502805857864616e-07, "loss": 0.0176, "step": 18820 }, { "grad_norm": 0.17936334013938904, "learning_rate": 9.34306000944396e-07, "loss": 0.013, "step": 18830 }, { "grad_norm": 0.16242825984954834, "learning_rate": 9.184655598361624e-07, "loss": 0.0132, "step": 18840 }, { "grad_norm": 0.12326671928167343, "learning_rate": 9.027593057689076e-07, "loss": 0.014, "step": 18850 }, { "grad_norm": 0.18215082585811615, "learning_rate": 8.871872816829441e-07, "loss": 0.0163, "step": 18860 }, { "grad_norm": 0.32729554176330566, "learning_rate": 8.717495301515777e-07, "loss": 0.0181, "step": 18870 }, { "grad_norm": 0.11798762530088425, "learning_rate": 8.564460933810415e-07, "loss": 0.0115, "step": 18880 }, { "grad_norm": 0.4078940451145172, "learning_rate": 8.412770132103453e-07, "loss": 0.0096, "step": 18890 }, { "grad_norm": 0.2021404355764389, "learning_rate": 8.262423311111711e-07, "loss": 0.0275, "step": 18900 }, { "grad_norm": 0.14557456970214844, "learning_rate": 8.113420881877665e-07, "loss": 0.0102, "step": 18910 }, { "grad_norm": 0.30091410875320435, "learning_rate": 7.965763251768288e-07, "loss": 0.0136, "step": 18920 }, { "grad_norm": 0.23842975497245789, "learning_rate": 7.819450824473995e-07, "loss": 0.0181, "step": 18930 }, { "grad_norm": 0.22350426018238068, "learning_rate": 7.674484000007198e-07, "loss": 0.0158, "step": 18940 }, { "grad_norm": 0.11302459985017776, "learning_rate": 7.530863174701752e-07, "loss": 0.0139, "step": 18950 }, { "grad_norm": 0.1941596269607544, "learning_rate": 7.38858874121151e-07, "loss": 0.0257, "step": 18960 }, { "grad_norm": 0.22360022366046906, "learning_rate": 7.247661088509328e-07, "loss": 0.0167, "step": 18970 }, { "grad_norm": 0.2868039906024933, "learning_rate": 7.108080601886002e-07, "loss": 0.0131, "step": 18980 }, { "grad_norm": 0.27363213896751404, "learning_rate": 6.969847662949336e-07, "loss": 0.0127, "step": 18990 }, { "grad_norm": 0.43462255597114563, "learning_rate": 6.832962649622798e-07, "loss": 0.0244, "step": 19000 }, { "grad_norm": 0.22143280506134033, "learning_rate": 6.697425936144863e-07, "loss": 0.0214, "step": 19010 }, { "grad_norm": 0.22814077138900757, "learning_rate": 6.563237893067731e-07, "loss": 0.0142, "step": 19020 }, { "grad_norm": 0.16559267044067383, "learning_rate": 6.430398887256328e-07, "loss": 0.0133, "step": 19030 }, { "grad_norm": 0.10037682950496674, "learning_rate": 6.298909281887478e-07, "loss": 0.0175, "step": 19040 }, { "grad_norm": 0.15412423014640808, "learning_rate": 6.168769436448673e-07, "loss": 0.0122, "step": 19050 }, { "grad_norm": 0.3349679708480835, "learning_rate": 6.03997970673742e-07, "loss": 0.019, "step": 19060 }, { "grad_norm": 0.18272764980793, "learning_rate": 5.912540444859782e-07, "loss": 0.0234, "step": 19070 }, { "grad_norm": 0.2876517176628113, "learning_rate": 5.786451999229837e-07, "loss": 0.0103, "step": 19080 }, { "grad_norm": 0.14135617017745972, "learning_rate": 5.661714714568722e-07, "loss": 0.0205, "step": 19090 }, { "grad_norm": 0.2004474550485611, "learning_rate": 5.538328931903259e-07, "loss": 0.0263, "step": 19100 }, { "grad_norm": 0.1586044430732727, "learning_rate": 5.416294988565551e-07, "loss": 0.018, "step": 19110 }, { "grad_norm": 0.14993205666542053, "learning_rate": 5.29561321819172e-07, "loss": 0.014, "step": 19120 }, { "grad_norm": 0.1110503226518631, "learning_rate": 5.176283950721061e-07, "loss": 0.0183, "step": 19130 }, { "grad_norm": 0.16163569688796997, "learning_rate": 5.058307512395332e-07, "loss": 0.0148, "step": 19140 }, { "grad_norm": 0.08035849034786224, "learning_rate": 4.941684225757526e-07, "loss": 0.0096, "step": 19150 }, { "grad_norm": 0.106717549264431, "learning_rate": 4.826414409651314e-07, "loss": 0.0141, "step": 19160 }, { "grad_norm": 0.3793148398399353, "learning_rate": 4.712498379219943e-07, "loss": 0.0155, "step": 19170 }, { "grad_norm": 0.19532422721385956, "learning_rate": 4.599936445905506e-07, "loss": 0.0151, "step": 19180 }, { "grad_norm": 0.29783132672309875, "learning_rate": 4.4887289174480594e-07, "loss": 0.0314, "step": 19190 }, { "grad_norm": 0.2437056303024292, "learning_rate": 4.378876097884621e-07, "loss": 0.0158, "step": 19200 }, { "grad_norm": 0.11475815623998642, "learning_rate": 4.2703782875487264e-07, "loss": 0.0169, "step": 19210 }, { "grad_norm": 0.34110400080680847, "learning_rate": 4.163235783069208e-07, "loss": 0.0188, "step": 19220 }, { "grad_norm": 0.1892102211713791, "learning_rate": 4.057448877369585e-07, "loss": 0.0193, "step": 19230 }, { "grad_norm": 0.12476175278425217, "learning_rate": 3.9530178596672295e-07, "loss": 0.0154, "step": 19240 }, { "grad_norm": 0.1520771086215973, "learning_rate": 3.849943015472479e-07, "loss": 0.0185, "step": 19250 }, { "grad_norm": 0.11702971160411835, "learning_rate": 3.748224626588137e-07, "loss": 0.0145, "step": 19260 }, { "grad_norm": 0.1393822878599167, "learning_rate": 3.647862971108307e-07, "loss": 0.0133, "step": 19270 }, { "grad_norm": 0.11174777895212173, "learning_rate": 3.5488583234179473e-07, "loss": 0.014, "step": 19280 }, { "grad_norm": 0.25525763630867004, "learning_rate": 3.4512109541920413e-07, "loss": 0.022, "step": 19290 }, { "grad_norm": 0.23591163754463196, "learning_rate": 3.354921130394706e-07, "loss": 0.015, "step": 19300 }, { "grad_norm": 0.3221985399723053, "learning_rate": 3.259989115278639e-07, "loss": 0.0146, "step": 19310 }, { "grad_norm": 0.13531263172626495, "learning_rate": 3.1664151683843403e-07, "loss": 0.0175, "step": 19320 }, { "grad_norm": 0.3794233500957489, "learning_rate": 3.074199545539447e-07, "loss": 0.0356, "step": 19330 }, { "grad_norm": 0.4987913966178894, "learning_rate": 2.983342498857955e-07, "loss": 0.0211, "step": 19340 }, { "grad_norm": 0.12399119883775711, "learning_rate": 2.893844276739499e-07, "loss": 0.0178, "step": 19350 }, { "grad_norm": 0.18568314611911774, "learning_rate": 2.8057051238688514e-07, "loss": 0.0144, "step": 19360 }, { "grad_norm": 0.24617396295070648, "learning_rate": 2.71892528121509e-07, "loss": 0.01, "step": 19370 }, { "grad_norm": 0.597692608833313, "learning_rate": 2.633504986030988e-07, "loss": 0.0253, "step": 19380 }, { "grad_norm": 0.052357934415340424, "learning_rate": 2.549444471852347e-07, "loss": 0.0125, "step": 19390 }, { "grad_norm": 0.43098098039627075, "learning_rate": 2.4667439684974423e-07, "loss": 0.0148, "step": 19400 }, { "grad_norm": 0.2889901399612427, "learning_rate": 2.3854037020662467e-07, "loss": 0.012, "step": 19410 }, { "grad_norm": 0.1694323569536209, "learning_rate": 2.3054238949399288e-07, "loss": 0.022, "step": 19420 }, { "grad_norm": 0.4363921880722046, "learning_rate": 2.2268047657802993e-07, "loss": 0.0203, "step": 19430 }, { "grad_norm": 0.1835341602563858, "learning_rate": 2.149546529529034e-07, "loss": 0.0173, "step": 19440 }, { "grad_norm": 0.2938174903392792, "learning_rate": 2.0736493974071736e-07, "loss": 0.0113, "step": 19450 }, { "grad_norm": 0.1998613476753235, "learning_rate": 1.9991135769145686e-07, "loss": 0.0231, "step": 19460 }, { "grad_norm": 0.1303967833518982, "learning_rate": 1.9259392718293245e-07, "loss": 0.0222, "step": 19470 }, { "grad_norm": 0.25173452496528625, "learning_rate": 1.8541266822072467e-07, "loss": 0.0163, "step": 19480 }, { "grad_norm": 0.17438781261444092, "learning_rate": 1.7836760043811184e-07, "loss": 0.0201, "step": 19490 }, { "grad_norm": 0.2208128720521927, "learning_rate": 1.7145874309604792e-07, "loss": 0.0148, "step": 19500 }, { "grad_norm": 0.14506326615810394, "learning_rate": 1.6468611508308474e-07, "loss": 0.0263, "step": 19510 }, { "grad_norm": 0.21628150343894958, "learning_rate": 1.5804973491532204e-07, "loss": 0.0108, "step": 19520 }, { "grad_norm": 0.16429480910301208, "learning_rate": 1.5154962073637424e-07, "loss": 0.0144, "step": 19530 }, { "grad_norm": 0.13973045349121094, "learning_rate": 1.4518579031730372e-07, "loss": 0.0152, "step": 19540 }, { "grad_norm": 0.37060776352882385, "learning_rate": 1.389582610565876e-07, "loss": 0.0206, "step": 19550 }, { "grad_norm": 0.19173160195350647, "learning_rate": 1.3286704998003995e-07, "loss": 0.0255, "step": 19560 }, { "grad_norm": 0.23848871886730194, "learning_rate": 1.2691217374080632e-07, "loss": 0.0214, "step": 19570 }, { "grad_norm": 0.12230383604764938, "learning_rate": 1.2109364861929705e-07, "loss": 0.0173, "step": 19580 }, { "grad_norm": 0.08463183790445328, "learning_rate": 1.1541149052312628e-07, "loss": 0.0176, "step": 19590 }, { "grad_norm": 0.24173837900161743, "learning_rate": 1.0986571498710074e-07, "loss": 0.0165, "step": 19600 }, { "grad_norm": 0.2703764736652374, "learning_rate": 1.0445633717316438e-07, "loss": 0.0149, "step": 19610 }, { "grad_norm": 0.1583162546157837, "learning_rate": 9.918337187034277e-08, "loss": 0.0138, "step": 19620 }, { "grad_norm": 0.03532618284225464, "learning_rate": 9.404683349472643e-08, "loss": 0.0133, "step": 19630 }, { "grad_norm": 0.09202857315540314, "learning_rate": 8.904673608940983e-08, "loss": 0.0127, "step": 19640 }, { "grad_norm": 0.24359239637851715, "learning_rate": 8.418309332447471e-08, "loss": 0.0301, "step": 19650 }, { "grad_norm": 0.3704041838645935, "learning_rate": 7.945591849692902e-08, "loss": 0.0136, "step": 19660 }, { "grad_norm": 0.06382566690444946, "learning_rate": 7.486522453069578e-08, "loss": 0.016, "step": 19670 }, { "grad_norm": 0.20399010181427002, "learning_rate": 7.041102397655208e-08, "loss": 0.0185, "step": 19680 }, { "grad_norm": 0.1005408763885498, "learning_rate": 6.609332901210685e-08, "loss": 0.0204, "step": 19690 }, { "grad_norm": 0.08737164735794067, "learning_rate": 6.191215144178419e-08, "loss": 0.0113, "step": 19700 }, { "grad_norm": 0.05895116552710533, "learning_rate": 5.786750269675678e-08, "loss": 0.016, "step": 19710 }, { "grad_norm": 0.4083814024925232, "learning_rate": 5.395939383494031e-08, "loss": 0.0198, "step": 19720 }, { "grad_norm": 0.30474188923835754, "learning_rate": 5.018783554095463e-08, "loss": 0.0156, "step": 19730 }, { "grad_norm": 0.21588455140590668, "learning_rate": 4.655283812610156e-08, "loss": 0.0168, "step": 19740 }, { "grad_norm": 0.20832733809947968, "learning_rate": 4.305441152831491e-08, "loss": 0.0109, "step": 19750 }, { "grad_norm": 0.15620902180671692, "learning_rate": 3.9692565312171584e-08, "loss": 0.0198, "step": 19760 }, { "grad_norm": 0.17725330591201782, "learning_rate": 3.6467308668824975e-08, "loss": 0.0094, "step": 19770 }, { "grad_norm": 0.24295781552791595, "learning_rate": 3.3378650416004964e-08, "loss": 0.0169, "step": 19780 }, { "grad_norm": 0.37783581018447876, "learning_rate": 3.042659899797906e-08, "loss": 0.0222, "step": 19790 }, { "grad_norm": 0.2539103031158447, "learning_rate": 2.76111624855524e-08, "loss": 0.016, "step": 19800 }, { "grad_norm": 0.4600794315338135, "learning_rate": 2.4932348576017784e-08, "loss": 0.0185, "step": 19810 }, { "grad_norm": 0.17680726945400238, "learning_rate": 2.239016459314458e-08, "loss": 0.0164, "step": 19820 }, { "grad_norm": 0.07962935417890549, "learning_rate": 1.9984617487173174e-08, "loss": 0.0107, "step": 19830 }, { "grad_norm": 0.16092652082443237, "learning_rate": 1.7715713834776105e-08, "loss": 0.0124, "step": 19840 }, { "grad_norm": 0.44818583130836487, "learning_rate": 1.5583459839046964e-08, "loss": 0.0208, "step": 19850 }, { "grad_norm": 0.22887055575847626, "learning_rate": 1.3587861329489304e-08, "loss": 0.0197, "step": 19860 }, { "grad_norm": 0.22888030111789703, "learning_rate": 1.1728923761994415e-08, "loss": 0.0178, "step": 19870 }, { "grad_norm": 0.20177464187145233, "learning_rate": 1.0006652218819135e-08, "loss": 0.0233, "step": 19880 }, { "grad_norm": 0.14045952260494232, "learning_rate": 8.421051408596947e-09, "loss": 0.0154, "step": 19890 }, { "grad_norm": 0.5360730886459351, "learning_rate": 6.972125666299123e-09, "loss": 0.0389, "step": 19900 }, { "grad_norm": 0.2244715839624405, "learning_rate": 5.659878953229169e-09, "loss": 0.0183, "step": 19910 }, { "grad_norm": 0.07774797081947327, "learning_rate": 4.48431485701728e-09, "loss": 0.0134, "step": 19920 }, { "grad_norm": 0.6350997686386108, "learning_rate": 3.4454365916203322e-09, "loss": 0.0159, "step": 19930 }, { "grad_norm": 0.23878128826618195, "learning_rate": 2.5432469972830332e-09, "loss": 0.0194, "step": 19940 }, { "grad_norm": 0.19011680781841278, "learning_rate": 1.7777485405601203e-09, "loss": 0.0118, "step": 19950 }, { "grad_norm": 0.2643686830997467, "learning_rate": 1.1489433142941597e-09, "loss": 0.0186, "step": 19960 }, { "grad_norm": 0.16888926923274994, "learning_rate": 6.568330376210963e-10, "loss": 0.0196, "step": 19970 }, { "grad_norm": 0.25553956627845764, "learning_rate": 3.0141905594249787e-10, "loss": 0.0178, "step": 19980 }, { "grad_norm": 0.16395509243011475, "learning_rate": 8.270234094776008e-11, "loss": 0.0168, "step": 19990 }, { "grad_norm": 0.13939934968948364, "learning_rate": 6.834906085551041e-13, "loss": 0.0161, "step": 20000 }, { "grad_norm": 0.2108602076768875, "learning_rate": 9.39666029309833e-05, "loss": 0.0316, "step": 20010 }, { "grad_norm": 1.228377103805542, "learning_rate": 9.395872655291596e-05, "loss": 0.0524, "step": 20020 }, { "grad_norm": 0.7210745215415955, "learning_rate": 9.395084536758838e-05, "loss": 0.044, "step": 20030 }, { "grad_norm": 0.5771674513816833, "learning_rate": 9.394295937586243e-05, "loss": 0.0439, "step": 20040 }, { "grad_norm": 0.3476361036300659, "learning_rate": 9.393506857860052e-05, "loss": 0.0622, "step": 20050 }, { "grad_norm": 0.5826542377471924, "learning_rate": 9.392717297666555e-05, "loss": 0.0377, "step": 20060 }, { "grad_norm": 0.6173798441886902, "learning_rate": 9.391927257092101e-05, "loss": 0.0553, "step": 20070 }, { "grad_norm": 0.28441208600997925, "learning_rate": 9.391136736223085e-05, "loss": 0.0445, "step": 20080 }, { "grad_norm": 0.3405860364437103, "learning_rate": 9.390345735145956e-05, "loss": 0.0425, "step": 20090 }, { "grad_norm": 0.3398333489894867, "learning_rate": 9.389554253947219e-05, "loss": 0.0435, "step": 20100 }, { "grad_norm": 0.22867351770401, "learning_rate": 9.388762292713428e-05, "loss": 0.0323, "step": 20110 }, { "grad_norm": 0.3586465120315552, "learning_rate": 9.38796985153119e-05, "loss": 0.0422, "step": 20120 }, { "grad_norm": 0.2984255850315094, "learning_rate": 9.387176930487169e-05, "loss": 0.0449, "step": 20130 }, { "grad_norm": 0.33211764693260193, "learning_rate": 9.386383529668072e-05, "loss": 0.0305, "step": 20140 }, { "grad_norm": 0.24771422147750854, "learning_rate": 9.385589649160669e-05, "loss": 0.0338, "step": 20150 }, { "grad_norm": 0.37046587467193604, "learning_rate": 9.384795289051775e-05, "loss": 0.0512, "step": 20160 }, { "grad_norm": 0.556082010269165, "learning_rate": 9.384000449428261e-05, "loss": 0.0463, "step": 20170 }, { "grad_norm": 0.5221240520477295, "learning_rate": 9.383205130377048e-05, "loss": 0.0571, "step": 20180 }, { "grad_norm": 0.3741579055786133, "learning_rate": 9.382409331985114e-05, "loss": 0.0396, "step": 20190 }, { "grad_norm": 0.27020159363746643, "learning_rate": 9.381613054339482e-05, "loss": 0.0336, "step": 20200 }, { "grad_norm": 0.2758673429489136, "learning_rate": 9.380816297527235e-05, "loss": 0.0397, "step": 20210 }, { "grad_norm": 0.28722622990608215, "learning_rate": 9.380019061635506e-05, "loss": 0.0342, "step": 20220 }, { "grad_norm": 0.42381584644317627, "learning_rate": 9.379221346751474e-05, "loss": 0.0559, "step": 20230 }, { "grad_norm": 0.6989412903785706, "learning_rate": 9.378423152962382e-05, "loss": 0.0427, "step": 20240 }, { "grad_norm": 0.31596627831459045, "learning_rate": 9.377624480355517e-05, "loss": 0.0367, "step": 20250 }, { "grad_norm": 0.30996963381767273, "learning_rate": 9.376825329018219e-05, "loss": 0.0347, "step": 20260 }, { "grad_norm": 0.2844100296497345, "learning_rate": 9.376025699037884e-05, "loss": 0.0352, "step": 20270 }, { "grad_norm": 0.3100437819957733, "learning_rate": 9.37522559050196e-05, "loss": 0.0343, "step": 20280 }, { "grad_norm": 0.2629917860031128, "learning_rate": 9.37442500349794e-05, "loss": 0.0533, "step": 20290 }, { "grad_norm": 0.24281233549118042, "learning_rate": 9.373623938113381e-05, "loss": 0.0341, "step": 20300 }, { "grad_norm": 0.2991427183151245, "learning_rate": 9.372822394435883e-05, "loss": 0.0438, "step": 20310 }, { "grad_norm": 0.31210774183273315, "learning_rate": 9.372020372553102e-05, "loss": 0.0278, "step": 20320 }, { "grad_norm": 0.6254821419715881, "learning_rate": 9.371217872552746e-05, "loss": 0.0442, "step": 20330 }, { "grad_norm": 0.34740617871284485, "learning_rate": 9.370414894522576e-05, "loss": 0.0362, "step": 20340 }, { "grad_norm": 0.48823297023773193, "learning_rate": 9.369611438550406e-05, "loss": 0.0395, "step": 20350 }, { "grad_norm": 0.3344736397266388, "learning_rate": 9.368807504724095e-05, "loss": 0.0439, "step": 20360 }, { "grad_norm": 0.5789849162101746, "learning_rate": 9.368003093131565e-05, "loss": 0.0489, "step": 20370 }, { "grad_norm": 0.4484966993331909, "learning_rate": 9.367198203860785e-05, "loss": 0.0787, "step": 20380 }, { "grad_norm": 2.6988446712493896, "learning_rate": 9.366392836999774e-05, "loss": 0.0507, "step": 20390 }, { "grad_norm": 0.2688528299331665, "learning_rate": 9.365586992636607e-05, "loss": 0.045, "step": 20400 }, { "grad_norm": 0.45628389716148376, "learning_rate": 9.364780670859412e-05, "loss": 0.0436, "step": 20410 }, { "grad_norm": 0.22291623055934906, "learning_rate": 9.363973871756364e-05, "loss": 0.0426, "step": 20420 }, { "grad_norm": 0.43834322690963745, "learning_rate": 9.363166595415696e-05, "loss": 0.036, "step": 20430 }, { "grad_norm": 0.3661963641643524, "learning_rate": 9.362358841925686e-05, "loss": 0.0466, "step": 20440 }, { "grad_norm": 0.5443698763847351, "learning_rate": 9.361550611374674e-05, "loss": 0.0501, "step": 20450 }, { "grad_norm": 0.3033737540245056, "learning_rate": 9.360741903851043e-05, "loss": 0.0626, "step": 20460 }, { "grad_norm": 0.3602355420589447, "learning_rate": 9.359932719443236e-05, "loss": 0.0515, "step": 20470 }, { "grad_norm": 0.7994487285614014, "learning_rate": 9.35912305823974e-05, "loss": 0.0462, "step": 20480 }, { "grad_norm": 1.3085092306137085, "learning_rate": 9.358312920329101e-05, "loss": 0.0608, "step": 20490 }, { "grad_norm": 0.47216665744781494, "learning_rate": 9.357502305799914e-05, "loss": 0.051, "step": 20500 }, { "grad_norm": 0.3050489127635956, "learning_rate": 9.356691214740824e-05, "loss": 0.0434, "step": 20510 }, { "grad_norm": 0.3076409101486206, "learning_rate": 9.355879647240535e-05, "loss": 0.0451, "step": 20520 }, { "grad_norm": 0.40500345826148987, "learning_rate": 9.355067603387798e-05, "loss": 0.042, "step": 20530 }, { "grad_norm": 0.5027915835380554, "learning_rate": 9.354255083271412e-05, "loss": 0.0355, "step": 20540 }, { "grad_norm": 0.3265702426433563, "learning_rate": 9.353442086980239e-05, "loss": 0.043, "step": 20550 }, { "grad_norm": 0.6801772713661194, "learning_rate": 9.352628614603185e-05, "loss": 0.0489, "step": 20560 }, { "grad_norm": 0.335562527179718, "learning_rate": 9.351814666229209e-05, "loss": 0.0421, "step": 20570 }, { "grad_norm": 0.4398689568042755, "learning_rate": 9.351000241947324e-05, "loss": 0.0355, "step": 20580 }, { "grad_norm": 0.3663155436515808, "learning_rate": 9.350185341846594e-05, "loss": 0.0394, "step": 20590 }, { "grad_norm": 0.23766405880451202, "learning_rate": 9.349369966016134e-05, "loss": 0.0551, "step": 20600 }, { "grad_norm": 0.4627220034599304, "learning_rate": 9.348554114545117e-05, "loss": 0.0595, "step": 20610 }, { "grad_norm": 0.3175952136516571, "learning_rate": 9.347737787522758e-05, "loss": 0.0517, "step": 20620 }, { "grad_norm": 0.3161259591579437, "learning_rate": 9.346920985038332e-05, "loss": 0.0416, "step": 20630 }, { "grad_norm": 0.4442351460456848, "learning_rate": 9.346103707181162e-05, "loss": 0.0398, "step": 20640 }, { "grad_norm": 0.33328282833099365, "learning_rate": 9.345285954040626e-05, "loss": 0.0311, "step": 20650 }, { "grad_norm": 0.49769559502601624, "learning_rate": 9.34446772570615e-05, "loss": 0.0417, "step": 20660 }, { "grad_norm": 0.34575706720352173, "learning_rate": 9.343649022267214e-05, "loss": 0.0329, "step": 20670 }, { "grad_norm": 0.361833393573761, "learning_rate": 9.342829843813353e-05, "loss": 0.0336, "step": 20680 }, { "grad_norm": 0.33128365874290466, "learning_rate": 9.342010190434149e-05, "loss": 0.0337, "step": 20690 }, { "grad_norm": 0.3843158185482025, "learning_rate": 9.34119006221924e-05, "loss": 0.0414, "step": 20700 }, { "grad_norm": 0.34638863801956177, "learning_rate": 9.340369459258313e-05, "loss": 0.0395, "step": 20710 }, { "grad_norm": 0.3485071063041687, "learning_rate": 9.339548381641106e-05, "loss": 0.0512, "step": 20720 }, { "grad_norm": 0.4265621602535248, "learning_rate": 9.338726829457413e-05, "loss": 0.0495, "step": 20730 }, { "grad_norm": 0.6065084934234619, "learning_rate": 9.337904802797078e-05, "loss": 0.0335, "step": 20740 }, { "grad_norm": 0.3392066955566406, "learning_rate": 9.337082301749993e-05, "loss": 0.0478, "step": 20750 }, { "grad_norm": 0.455527126789093, "learning_rate": 9.336259326406109e-05, "loss": 0.0405, "step": 20760 }, { "grad_norm": 0.3614440858364105, "learning_rate": 9.335435876855427e-05, "loss": 0.0566, "step": 20770 }, { "grad_norm": 0.3313816785812378, "learning_rate": 9.334611953187994e-05, "loss": 0.0498, "step": 20780 }, { "grad_norm": 0.5267049670219421, "learning_rate": 9.333787555493914e-05, "loss": 0.0547, "step": 20790 }, { "grad_norm": 0.5088739395141602, "learning_rate": 9.332962683863345e-05, "loss": 0.0484, "step": 20800 }, { "grad_norm": 0.2590878903865814, "learning_rate": 9.332137338386489e-05, "loss": 0.0388, "step": 20810 }, { "grad_norm": 0.7079911828041077, "learning_rate": 9.33131151915361e-05, "loss": 0.0577, "step": 20820 }, { "grad_norm": 0.3511272668838501, "learning_rate": 9.330485226255012e-05, "loss": 0.0332, "step": 20830 }, { "grad_norm": 0.3814012110233307, "learning_rate": 9.329658459781061e-05, "loss": 0.0466, "step": 20840 }, { "grad_norm": 0.36364907026290894, "learning_rate": 9.328831219822172e-05, "loss": 0.0353, "step": 20850 }, { "grad_norm": 0.30995944142341614, "learning_rate": 9.328003506468808e-05, "loss": 0.0365, "step": 20860 }, { "grad_norm": 0.4662267863750458, "learning_rate": 9.327175319811488e-05, "loss": 0.0378, "step": 20870 }, { "grad_norm": 0.3281327784061432, "learning_rate": 9.326346659940781e-05, "loss": 0.0326, "step": 20880 }, { "grad_norm": 0.39018911123275757, "learning_rate": 9.325517526947308e-05, "loss": 0.0437, "step": 20890 }, { "grad_norm": 0.5898265242576599, "learning_rate": 9.32468792092174e-05, "loss": 0.0417, "step": 20900 }, { "grad_norm": 0.3307174742221832, "learning_rate": 9.323857841954803e-05, "loss": 0.0444, "step": 20910 }, { "grad_norm": 0.2710389792919159, "learning_rate": 9.323027290137276e-05, "loss": 0.036, "step": 20920 }, { "grad_norm": 0.31860584020614624, "learning_rate": 9.322196265559981e-05, "loss": 0.0425, "step": 20930 }, { "grad_norm": 0.23526400327682495, "learning_rate": 9.321364768313803e-05, "loss": 0.0456, "step": 20940 }, { "grad_norm": 0.4131414592266083, "learning_rate": 9.32053279848967e-05, "loss": 0.0481, "step": 20950 }, { "grad_norm": 0.627827525138855, "learning_rate": 9.319700356178567e-05, "loss": 0.0407, "step": 20960 }, { "grad_norm": 0.3247920870780945, "learning_rate": 9.318867441471527e-05, "loss": 0.0397, "step": 20970 }, { "grad_norm": 0.3265015184879303, "learning_rate": 9.318034054459637e-05, "loss": 0.033, "step": 20980 }, { "grad_norm": 0.39705535769462585, "learning_rate": 9.317200195234034e-05, "loss": 0.0329, "step": 20990 }, { "grad_norm": 0.275162935256958, "learning_rate": 9.316365863885909e-05, "loss": 0.0439, "step": 21000 }, { "grad_norm": 0.4601873755455017, "learning_rate": 9.315531060506502e-05, "loss": 0.0494, "step": 21010 }, { "grad_norm": 0.279399037361145, "learning_rate": 9.314695785187108e-05, "loss": 0.0374, "step": 21020 }, { "grad_norm": 0.4652799069881439, "learning_rate": 9.313860038019069e-05, "loss": 0.0454, "step": 21030 }, { "grad_norm": 0.2279638797044754, "learning_rate": 9.313023819093782e-05, "loss": 0.0412, "step": 21040 }, { "grad_norm": 0.3310677111148834, "learning_rate": 9.312187128502695e-05, "loss": 0.0377, "step": 21050 }, { "grad_norm": 0.704784095287323, "learning_rate": 9.311349966337307e-05, "loss": 0.0564, "step": 21060 }, { "grad_norm": 0.4658893048763275, "learning_rate": 9.310512332689169e-05, "loss": 0.0365, "step": 21070 }, { "grad_norm": 0.3269941806793213, "learning_rate": 9.309674227649883e-05, "loss": 0.0389, "step": 21080 }, { "grad_norm": 0.7156462669372559, "learning_rate": 9.308835651311103e-05, "loss": 0.061, "step": 21090 }, { "grad_norm": 0.45336559414863586, "learning_rate": 9.307996603764533e-05, "loss": 0.0491, "step": 21100 }, { "grad_norm": 0.390847384929657, "learning_rate": 9.307157085101932e-05, "loss": 0.0355, "step": 21110 }, { "grad_norm": 0.30821844935417175, "learning_rate": 9.306317095415109e-05, "loss": 0.0337, "step": 21120 }, { "grad_norm": 0.3969907760620117, "learning_rate": 9.305476634795922e-05, "loss": 0.0516, "step": 21130 }, { "grad_norm": 0.3948042094707489, "learning_rate": 9.304635703336284e-05, "loss": 0.0443, "step": 21140 }, { "grad_norm": 0.2944413423538208, "learning_rate": 9.303794301128157e-05, "loss": 0.0367, "step": 21150 }, { "grad_norm": 0.39140138030052185, "learning_rate": 9.302952428263555e-05, "loss": 0.0297, "step": 21160 }, { "grad_norm": 0.3227570652961731, "learning_rate": 9.302110084834545e-05, "loss": 0.0328, "step": 21170 }, { "grad_norm": 0.32633283734321594, "learning_rate": 9.301267270933245e-05, "loss": 0.0336, "step": 21180 }, { "grad_norm": 0.301556795835495, "learning_rate": 9.300423986651823e-05, "loss": 0.0403, "step": 21190 }, { "grad_norm": 0.5305954813957214, "learning_rate": 9.299580232082501e-05, "loss": 0.0438, "step": 21200 }, { "grad_norm": 0.6097064018249512, "learning_rate": 9.298736007317547e-05, "loss": 0.0412, "step": 21210 }, { "grad_norm": 0.4755305349826813, "learning_rate": 9.297891312449288e-05, "loss": 0.0543, "step": 21220 }, { "grad_norm": 0.5991378426551819, "learning_rate": 9.297046147570094e-05, "loss": 0.0414, "step": 21230 }, { "grad_norm": 0.2999233901500702, "learning_rate": 9.296200512772396e-05, "loss": 0.0283, "step": 21240 }, { "grad_norm": 0.29982295632362366, "learning_rate": 9.295354408148668e-05, "loss": 0.0332, "step": 21250 }, { "grad_norm": 0.38693615794181824, "learning_rate": 9.294507833791441e-05, "loss": 0.0501, "step": 21260 }, { "grad_norm": 0.2675299644470215, "learning_rate": 9.293660789793295e-05, "loss": 0.0308, "step": 21270 }, { "grad_norm": 0.45677363872528076, "learning_rate": 9.292813276246858e-05, "loss": 0.0322, "step": 21280 }, { "grad_norm": 0.2885838449001312, "learning_rate": 9.291965293244816e-05, "loss": 0.0439, "step": 21290 }, { "grad_norm": 0.391430526971817, "learning_rate": 9.291116840879904e-05, "loss": 0.0325, "step": 21300 }, { "grad_norm": 0.3881066143512726, "learning_rate": 9.290267919244904e-05, "loss": 0.0323, "step": 21310 }, { "grad_norm": 0.17709973454475403, "learning_rate": 9.289418528432655e-05, "loss": 0.0232, "step": 21320 }, { "grad_norm": 0.24301788210868835, "learning_rate": 9.288568668536045e-05, "loss": 0.0468, "step": 21330 }, { "grad_norm": 0.46147510409355164, "learning_rate": 9.287718339648013e-05, "loss": 0.0371, "step": 21340 }, { "grad_norm": 0.4815654754638672, "learning_rate": 9.28686754186155e-05, "loss": 0.033, "step": 21350 }, { "grad_norm": 0.20510387420654297, "learning_rate": 9.286016275269698e-05, "loss": 0.0281, "step": 21360 }, { "grad_norm": 0.3611491918563843, "learning_rate": 9.285164539965551e-05, "loss": 0.0268, "step": 21370 }, { "grad_norm": 0.5260391235351562, "learning_rate": 9.284312336042251e-05, "loss": 0.0321, "step": 21380 }, { "grad_norm": 0.34656384587287903, "learning_rate": 9.283459663592996e-05, "loss": 0.0319, "step": 21390 }, { "grad_norm": 0.296629935503006, "learning_rate": 9.282606522711033e-05, "loss": 0.0385, "step": 21400 }, { "grad_norm": 0.19132699072360992, "learning_rate": 9.281752913489657e-05, "loss": 0.0499, "step": 21410 }, { "grad_norm": 0.23553474247455597, "learning_rate": 9.280898836022222e-05, "loss": 0.0388, "step": 21420 }, { "grad_norm": 0.4681957960128784, "learning_rate": 9.280044290402126e-05, "loss": 0.0278, "step": 21430 }, { "grad_norm": 0.2711832523345947, "learning_rate": 9.279189276722821e-05, "loss": 0.023, "step": 21440 }, { "grad_norm": 0.34786123037338257, "learning_rate": 9.278333795077812e-05, "loss": 0.0477, "step": 21450 }, { "grad_norm": 0.2891610860824585, "learning_rate": 9.27747784556065e-05, "loss": 0.0225, "step": 21460 }, { "grad_norm": 0.3745017945766449, "learning_rate": 9.276621428264942e-05, "loss": 0.0537, "step": 21470 }, { "grad_norm": 0.2530493140220642, "learning_rate": 9.275764543284345e-05, "loss": 0.0258, "step": 21480 }, { "grad_norm": 0.36276745796203613, "learning_rate": 9.274907190712566e-05, "loss": 0.0296, "step": 21490 }, { "grad_norm": 0.2945054769515991, "learning_rate": 9.274049370643363e-05, "loss": 0.0322, "step": 21500 }, { "grad_norm": 0.29781532287597656, "learning_rate": 9.273191083170547e-05, "loss": 0.0375, "step": 21510 }, { "grad_norm": 0.30789291858673096, "learning_rate": 9.27233232838798e-05, "loss": 0.0434, "step": 21520 }, { "grad_norm": 0.3597237765789032, "learning_rate": 9.27147310638957e-05, "loss": 0.0468, "step": 21530 }, { "grad_norm": 0.3415672183036804, "learning_rate": 9.270613417269286e-05, "loss": 0.0319, "step": 21540 }, { "grad_norm": 0.7196137309074402, "learning_rate": 9.269753261121138e-05, "loss": 0.0337, "step": 21550 }, { "grad_norm": 0.1918921172618866, "learning_rate": 9.268892638039194e-05, "loss": 0.0319, "step": 21560 }, { "grad_norm": 0.3686569035053253, "learning_rate": 9.268031548117569e-05, "loss": 0.035, "step": 21570 }, { "grad_norm": 0.3955176770687103, "learning_rate": 9.26716999145043e-05, "loss": 0.0342, "step": 21580 }, { "grad_norm": 0.37187469005584717, "learning_rate": 9.266307968131998e-05, "loss": 0.0406, "step": 21590 }, { "grad_norm": 0.3510473966598511, "learning_rate": 9.26544547825654e-05, "loss": 0.0301, "step": 21600 }, { "grad_norm": 0.3485584557056427, "learning_rate": 9.264582521918376e-05, "loss": 0.0453, "step": 21610 }, { "grad_norm": 0.27977004647254944, "learning_rate": 9.263719099211881e-05, "loss": 0.0306, "step": 21620 }, { "grad_norm": 0.2534559369087219, "learning_rate": 9.262855210231476e-05, "loss": 0.0348, "step": 21630 }, { "grad_norm": 0.376783162355423, "learning_rate": 9.261990855071633e-05, "loss": 0.038, "step": 21640 }, { "grad_norm": 0.3220979571342468, "learning_rate": 9.261126033826878e-05, "loss": 0.0312, "step": 21650 }, { "grad_norm": 0.3911856710910797, "learning_rate": 9.260260746591786e-05, "loss": 0.0487, "step": 21660 }, { "grad_norm": 0.22314117848873138, "learning_rate": 9.259394993460985e-05, "loss": 0.0315, "step": 21670 }, { "grad_norm": 0.20223698019981384, "learning_rate": 9.258528774529151e-05, "loss": 0.0305, "step": 21680 }, { "grad_norm": 0.3928617835044861, "learning_rate": 9.257662089891013e-05, "loss": 0.0442, "step": 21690 }, { "grad_norm": 0.3779287338256836, "learning_rate": 9.25679493964135e-05, "loss": 0.0328, "step": 21700 }, { "grad_norm": 0.3575919270515442, "learning_rate": 9.255927323874994e-05, "loss": 0.0323, "step": 21710 }, { "grad_norm": 0.2926061749458313, "learning_rate": 9.255059242686822e-05, "loss": 0.0311, "step": 21720 }, { "grad_norm": 0.3750928044319153, "learning_rate": 9.254190696171769e-05, "loss": 0.0279, "step": 21730 }, { "grad_norm": 0.28028154373168945, "learning_rate": 9.25332168442482e-05, "loss": 0.031, "step": 21740 }, { "grad_norm": 0.35291755199432373, "learning_rate": 9.252452207541004e-05, "loss": 0.037, "step": 21750 }, { "grad_norm": 0.4384055733680725, "learning_rate": 9.251582265615409e-05, "loss": 0.0394, "step": 21760 }, { "grad_norm": 0.46069878339767456, "learning_rate": 9.250711858743169e-05, "loss": 0.0378, "step": 21770 }, { "grad_norm": 0.4236273169517517, "learning_rate": 9.24984098701947e-05, "loss": 0.0452, "step": 21780 }, { "grad_norm": 0.37211647629737854, "learning_rate": 9.248969650539552e-05, "loss": 0.0317, "step": 21790 }, { "grad_norm": 0.28940773010253906, "learning_rate": 9.2480978493987e-05, "loss": 0.0385, "step": 21800 }, { "grad_norm": 0.2703116834163666, "learning_rate": 9.247225583692256e-05, "loss": 0.0312, "step": 21810 }, { "grad_norm": 0.32423993945121765, "learning_rate": 9.246352853515607e-05, "loss": 0.0336, "step": 21820 }, { "grad_norm": 0.4054764211177826, "learning_rate": 9.245479658964194e-05, "loss": 0.0397, "step": 21830 }, { "grad_norm": 0.4007608890533447, "learning_rate": 9.244606000133507e-05, "loss": 0.0369, "step": 21840 }, { "grad_norm": 0.31816354393959045, "learning_rate": 9.24373187711909e-05, "loss": 0.0277, "step": 21850 }, { "grad_norm": 0.33916816115379333, "learning_rate": 9.242857290016537e-05, "loss": 0.0433, "step": 21860 }, { "grad_norm": 0.5183128118515015, "learning_rate": 9.241982238921488e-05, "loss": 0.0417, "step": 21870 }, { "grad_norm": 0.40960052609443665, "learning_rate": 9.24110672392964e-05, "loss": 0.0332, "step": 21880 }, { "grad_norm": 0.32372716069221497, "learning_rate": 9.240230745136737e-05, "loss": 0.0363, "step": 21890 }, { "grad_norm": 0.35698258876800537, "learning_rate": 9.239354302638575e-05, "loss": 0.0384, "step": 21900 }, { "grad_norm": 0.255918025970459, "learning_rate": 9.238477396531e-05, "loss": 0.0333, "step": 21910 }, { "grad_norm": 0.25457608699798584, "learning_rate": 9.23760002690991e-05, "loss": 0.0388, "step": 21920 }, { "grad_norm": 0.3064086437225342, "learning_rate": 9.236722193871252e-05, "loss": 0.0384, "step": 21930 }, { "grad_norm": 0.46251314878463745, "learning_rate": 9.235843897511023e-05, "loss": 0.048, "step": 21940 }, { "grad_norm": 0.3597399890422821, "learning_rate": 9.234965137925276e-05, "loss": 0.029, "step": 21950 }, { "grad_norm": 0.31501051783561707, "learning_rate": 9.234085915210108e-05, "loss": 0.042, "step": 21960 }, { "grad_norm": 0.2415153533220291, "learning_rate": 9.23320622946167e-05, "loss": 0.0283, "step": 21970 }, { "grad_norm": 0.14801061153411865, "learning_rate": 9.232326080776163e-05, "loss": 0.0273, "step": 21980 }, { "grad_norm": 0.2783311605453491, "learning_rate": 9.23144546924984e-05, "loss": 0.0301, "step": 21990 }, { "grad_norm": 0.29189661145210266, "learning_rate": 9.230564394979e-05, "loss": 0.0373, "step": 22000 }, { "grad_norm": 0.4349606931209564, "learning_rate": 9.22968285806e-05, "loss": 0.0285, "step": 22010 }, { "grad_norm": 0.3284573256969452, "learning_rate": 9.228800858589242e-05, "loss": 0.0464, "step": 22020 }, { "grad_norm": 0.20125631988048553, "learning_rate": 9.227918396663179e-05, "loss": 0.0327, "step": 22030 }, { "grad_norm": 0.3610744774341583, "learning_rate": 9.227035472378319e-05, "loss": 0.045, "step": 22040 }, { "grad_norm": 0.20198528468608856, "learning_rate": 9.226152085831213e-05, "loss": 0.0315, "step": 22050 }, { "grad_norm": 0.28601762652397156, "learning_rate": 9.22526823711847e-05, "loss": 0.0292, "step": 22060 }, { "grad_norm": 0.11830522119998932, "learning_rate": 9.224383926336745e-05, "loss": 0.0357, "step": 22070 }, { "grad_norm": 0.4525901675224304, "learning_rate": 9.223499153582744e-05, "loss": 0.0247, "step": 22080 }, { "grad_norm": 0.4328157901763916, "learning_rate": 9.222613918953226e-05, "loss": 0.0474, "step": 22090 }, { "grad_norm": 0.25427359342575073, "learning_rate": 9.221728222544999e-05, "loss": 0.0368, "step": 22100 }, { "grad_norm": 0.5434119701385498, "learning_rate": 9.22084206445492e-05, "loss": 0.0512, "step": 22110 }, { "grad_norm": 0.2149258702993393, "learning_rate": 9.2199554447799e-05, "loss": 0.037, "step": 22120 }, { "grad_norm": 0.20332573354244232, "learning_rate": 9.219068363616897e-05, "loss": 0.0378, "step": 22130 }, { "grad_norm": 0.46965715289115906, "learning_rate": 9.218180821062919e-05, "loss": 0.0338, "step": 22140 }, { "grad_norm": 0.28676334023475647, "learning_rate": 9.21729281721503e-05, "loss": 0.0302, "step": 22150 }, { "grad_norm": 0.24342456459999084, "learning_rate": 9.216404352170339e-05, "loss": 0.0251, "step": 22160 }, { "grad_norm": 0.24400264024734497, "learning_rate": 9.215515426026007e-05, "loss": 0.0268, "step": 22170 }, { "grad_norm": 0.33935871720314026, "learning_rate": 9.214626038879246e-05, "loss": 0.0294, "step": 22180 }, { "grad_norm": 0.2683200240135193, "learning_rate": 9.21373619082732e-05, "loss": 0.0269, "step": 22190 }, { "grad_norm": 0.33537042140960693, "learning_rate": 9.212845881967535e-05, "loss": 0.048, "step": 22200 }, { "grad_norm": 0.5616001486778259, "learning_rate": 9.211955112397262e-05, "loss": 0.0363, "step": 22210 }, { "grad_norm": 0.4129147231578827, "learning_rate": 9.211063882213909e-05, "loss": 0.0398, "step": 22220 }, { "grad_norm": 0.2353297770023346, "learning_rate": 9.210172191514942e-05, "loss": 0.0302, "step": 22230 }, { "grad_norm": 0.41104352474212646, "learning_rate": 9.209280040397874e-05, "loss": 0.0444, "step": 22240 }, { "grad_norm": 0.3074972927570343, "learning_rate": 9.208387428960268e-05, "loss": 0.0289, "step": 22250 }, { "grad_norm": 0.46910303831100464, "learning_rate": 9.20749435729974e-05, "loss": 0.0318, "step": 22260 }, { "grad_norm": 0.4077117443084717, "learning_rate": 9.206600825513957e-05, "loss": 0.0268, "step": 22270 }, { "grad_norm": 0.20391950011253357, "learning_rate": 9.20570683370063e-05, "loss": 0.0345, "step": 22280 }, { "grad_norm": 0.5825697779655457, "learning_rate": 9.204812381957528e-05, "loss": 0.0282, "step": 22290 }, { "grad_norm": 0.5943038463592529, "learning_rate": 9.203917470382465e-05, "loss": 0.0374, "step": 22300 }, { "grad_norm": 0.3805738389492035, "learning_rate": 9.203022099073309e-05, "loss": 0.0445, "step": 22310 }, { "grad_norm": 0.2943059802055359, "learning_rate": 9.202126268127976e-05, "loss": 0.0391, "step": 22320 }, { "grad_norm": 0.312298983335495, "learning_rate": 9.20122997764443e-05, "loss": 0.0359, "step": 22330 }, { "grad_norm": 0.4563951790332794, "learning_rate": 9.200333227720692e-05, "loss": 0.0506, "step": 22340 }, { "grad_norm": 0.3361164331436157, "learning_rate": 9.199436018454826e-05, "loss": 0.0386, "step": 22350 }, { "grad_norm": 0.16662569344043732, "learning_rate": 9.198538349944952e-05, "loss": 0.0317, "step": 22360 }, { "grad_norm": 0.3862639367580414, "learning_rate": 9.197640222289234e-05, "loss": 0.0504, "step": 22370 }, { "grad_norm": 0.2972208261489868, "learning_rate": 9.196741635585895e-05, "loss": 0.0321, "step": 22380 }, { "grad_norm": 0.26370638608932495, "learning_rate": 9.195842589933199e-05, "loss": 0.0315, "step": 22390 }, { "grad_norm": 0.3086877167224884, "learning_rate": 9.194943085429466e-05, "loss": 0.0362, "step": 22400 }, { "grad_norm": 0.4620507061481476, "learning_rate": 9.194043122173065e-05, "loss": 0.0481, "step": 22410 }, { "grad_norm": 0.3589612543582916, "learning_rate": 9.193142700262413e-05, "loss": 0.048, "step": 22420 }, { "grad_norm": 0.3268122971057892, "learning_rate": 9.192241819795979e-05, "loss": 0.0359, "step": 22430 }, { "grad_norm": 0.32954367995262146, "learning_rate": 9.191340480872284e-05, "loss": 0.0365, "step": 22440 }, { "grad_norm": 0.3518420159816742, "learning_rate": 9.190438683589895e-05, "loss": 0.0345, "step": 22450 }, { "grad_norm": 0.35350748896598816, "learning_rate": 9.189536428047432e-05, "loss": 0.0354, "step": 22460 }, { "grad_norm": 0.31667110323905945, "learning_rate": 9.188633714343564e-05, "loss": 0.0409, "step": 22470 }, { "grad_norm": 0.8142223358154297, "learning_rate": 9.18773054257701e-05, "loss": 0.0425, "step": 22480 }, { "grad_norm": 0.39836347103118896, "learning_rate": 9.18682691284654e-05, "loss": 0.0386, "step": 22490 }, { "grad_norm": 0.3413175642490387, "learning_rate": 9.185922825250974e-05, "loss": 0.0415, "step": 22500 }, { "grad_norm": 0.159648597240448, "learning_rate": 9.185018279889181e-05, "loss": 0.0213, "step": 22510 }, { "grad_norm": 0.42207542061805725, "learning_rate": 9.184113276860082e-05, "loss": 0.0371, "step": 22520 }, { "grad_norm": 0.24547770619392395, "learning_rate": 9.183207816262645e-05, "loss": 0.0465, "step": 22530 }, { "grad_norm": 0.34508490562438965, "learning_rate": 9.182301898195891e-05, "loss": 0.025, "step": 22540 }, { "grad_norm": 0.1811836063861847, "learning_rate": 9.181395522758889e-05, "loss": 0.0374, "step": 22550 }, { "grad_norm": 0.40145546197891235, "learning_rate": 9.180488690050759e-05, "loss": 0.0432, "step": 22560 }, { "grad_norm": 0.2787530720233917, "learning_rate": 9.179581400170671e-05, "loss": 0.0294, "step": 22570 }, { "grad_norm": 0.2182088941335678, "learning_rate": 9.178673653217845e-05, "loss": 0.0348, "step": 22580 }, { "grad_norm": 0.3571494221687317, "learning_rate": 9.177765449291551e-05, "loss": 0.0277, "step": 22590 }, { "grad_norm": 0.1439325213432312, "learning_rate": 9.176856788491109e-05, "loss": 0.0276, "step": 22600 }, { "grad_norm": 0.32344672083854675, "learning_rate": 9.175947670915887e-05, "loss": 0.0319, "step": 22610 }, { "grad_norm": 0.2687189280986786, "learning_rate": 9.175038096665309e-05, "loss": 0.0326, "step": 22620 }, { "grad_norm": 0.508471667766571, "learning_rate": 9.17412806583884e-05, "loss": 0.0312, "step": 22630 }, { "grad_norm": 0.23471258580684662, "learning_rate": 9.173217578536002e-05, "loss": 0.0381, "step": 22640 }, { "grad_norm": 0.17607076466083527, "learning_rate": 9.172306634856362e-05, "loss": 0.0397, "step": 22650 }, { "grad_norm": 0.4505693018436432, "learning_rate": 9.171395234899545e-05, "loss": 0.0273, "step": 22660 }, { "grad_norm": 0.2670859396457672, "learning_rate": 9.170483378765214e-05, "loss": 0.0309, "step": 22670 }, { "grad_norm": 0.39294004440307617, "learning_rate": 9.169571066553091e-05, "loss": 0.0447, "step": 22680 }, { "grad_norm": 0.294622540473938, "learning_rate": 9.168658298362946e-05, "loss": 0.027, "step": 22690 }, { "grad_norm": 0.2130836844444275, "learning_rate": 9.167745074294598e-05, "loss": 0.0321, "step": 22700 }, { "grad_norm": 0.22664599120616913, "learning_rate": 9.166831394447913e-05, "loss": 0.0357, "step": 22710 }, { "grad_norm": 0.1875155121088028, "learning_rate": 9.165917258922812e-05, "loss": 0.0264, "step": 22720 }, { "grad_norm": 0.3308604657649994, "learning_rate": 9.165002667819262e-05, "loss": 0.0385, "step": 22730 }, { "grad_norm": 0.29773566126823425, "learning_rate": 9.164087621237282e-05, "loss": 0.0317, "step": 22740 }, { "grad_norm": 0.5479252934455872, "learning_rate": 9.163172119276942e-05, "loss": 0.0379, "step": 22750 }, { "grad_norm": 0.18782109022140503, "learning_rate": 9.162256162038358e-05, "loss": 0.044, "step": 22760 }, { "grad_norm": 0.24660657346248627, "learning_rate": 9.161339749621698e-05, "loss": 0.0318, "step": 22770 }, { "grad_norm": 0.3059741258621216, "learning_rate": 9.160422882127177e-05, "loss": 0.0329, "step": 22780 }, { "grad_norm": 0.2874390780925751, "learning_rate": 9.159505559655069e-05, "loss": 0.0201, "step": 22790 }, { "grad_norm": 0.2945641577243805, "learning_rate": 9.158587782305684e-05, "loss": 0.0273, "step": 22800 }, { "grad_norm": 0.3617236912250519, "learning_rate": 9.157669550179391e-05, "loss": 0.0297, "step": 22810 }, { "grad_norm": 0.3012509047985077, "learning_rate": 9.156750863376609e-05, "loss": 0.0404, "step": 22820 }, { "grad_norm": 0.3567028343677521, "learning_rate": 9.155831721997801e-05, "loss": 0.0301, "step": 22830 }, { "grad_norm": 0.3962401747703552, "learning_rate": 9.154912126143484e-05, "loss": 0.0415, "step": 22840 }, { "grad_norm": 0.4595274329185486, "learning_rate": 9.153992075914224e-05, "loss": 0.0512, "step": 22850 }, { "grad_norm": 0.39879316091537476, "learning_rate": 9.153071571410635e-05, "loss": 0.0313, "step": 22860 }, { "grad_norm": 0.27856332063674927, "learning_rate": 9.152150612733384e-05, "loss": 0.0392, "step": 22870 }, { "grad_norm": 0.2843668460845947, "learning_rate": 9.151229199983184e-05, "loss": 0.0367, "step": 22880 }, { "grad_norm": 0.37398532032966614, "learning_rate": 9.150307333260802e-05, "loss": 0.0425, "step": 22890 }, { "grad_norm": 0.5977532863616943, "learning_rate": 9.149385012667048e-05, "loss": 0.049, "step": 22900 }, { "grad_norm": 0.31529197096824646, "learning_rate": 9.148462238302788e-05, "loss": 0.0325, "step": 22910 }, { "grad_norm": 0.3656003475189209, "learning_rate": 9.147539010268936e-05, "loss": 0.0436, "step": 22920 }, { "grad_norm": 0.22250257432460785, "learning_rate": 9.14661532866645e-05, "loss": 0.0299, "step": 22930 }, { "grad_norm": 0.1509338766336441, "learning_rate": 9.145691193596348e-05, "loss": 0.025, "step": 22940 }, { "grad_norm": 0.30297568440437317, "learning_rate": 9.144766605159691e-05, "loss": 0.0405, "step": 22950 }, { "grad_norm": 0.27565962076187134, "learning_rate": 9.14384156345759e-05, "loss": 0.0302, "step": 22960 }, { "grad_norm": 0.2126595377922058, "learning_rate": 9.142916068591204e-05, "loss": 0.0227, "step": 22970 }, { "grad_norm": 0.32097530364990234, "learning_rate": 9.141990120661746e-05, "loss": 0.0347, "step": 22980 }, { "grad_norm": 0.22315570712089539, "learning_rate": 9.141063719770475e-05, "loss": 0.0271, "step": 22990 }, { "grad_norm": 0.38899943232536316, "learning_rate": 9.140136866018704e-05, "loss": 0.0291, "step": 23000 }, { "grad_norm": 0.2686997354030609, "learning_rate": 9.139209559507788e-05, "loss": 0.0273, "step": 23010 }, { "grad_norm": 0.32551446557044983, "learning_rate": 9.13828180033914e-05, "loss": 0.0263, "step": 23020 }, { "grad_norm": 0.2997841238975525, "learning_rate": 9.137353588614212e-05, "loss": 0.0386, "step": 23030 }, { "grad_norm": 0.33491599559783936, "learning_rate": 9.136424924434519e-05, "loss": 0.0303, "step": 23040 }, { "grad_norm": 0.6825210452079773, "learning_rate": 9.135495807901615e-05, "loss": 0.0412, "step": 23050 }, { "grad_norm": 0.2870030999183655, "learning_rate": 9.134566239117108e-05, "loss": 0.0316, "step": 23060 }, { "grad_norm": 0.21175113320350647, "learning_rate": 9.13363621818265e-05, "loss": 0.0302, "step": 23070 }, { "grad_norm": 0.3034599721431732, "learning_rate": 9.132705745199953e-05, "loss": 0.0215, "step": 23080 }, { "grad_norm": 0.30483320355415344, "learning_rate": 9.131774820270768e-05, "loss": 0.0309, "step": 23090 }, { "grad_norm": 0.22818031907081604, "learning_rate": 9.130843443496901e-05, "loss": 0.0375, "step": 23100 }, { "grad_norm": 0.2896588444709778, "learning_rate": 9.129911614980206e-05, "loss": 0.0431, "step": 23110 }, { "grad_norm": 0.17727714776992798, "learning_rate": 9.128979334822584e-05, "loss": 0.0424, "step": 23120 }, { "grad_norm": 0.25353536009788513, "learning_rate": 9.128046603125992e-05, "loss": 0.0255, "step": 23130 }, { "grad_norm": 0.1584252566099167, "learning_rate": 9.12711341999243e-05, "loss": 0.0229, "step": 23140 }, { "grad_norm": 0.2991119623184204, "learning_rate": 9.12617978552395e-05, "loss": 0.0495, "step": 23150 }, { "grad_norm": 0.19010914862155914, "learning_rate": 9.12524569982265e-05, "loss": 0.0287, "step": 23160 }, { "grad_norm": 0.23725400865077972, "learning_rate": 9.124311162990684e-05, "loss": 0.0329, "step": 23170 }, { "grad_norm": 0.48727983236312866, "learning_rate": 9.12337617513025e-05, "loss": 0.0287, "step": 23180 }, { "grad_norm": 0.3431796729564667, "learning_rate": 9.122440736343596e-05, "loss": 0.0285, "step": 23190 }, { "grad_norm": 0.32517850399017334, "learning_rate": 9.12150484673302e-05, "loss": 0.0295, "step": 23200 }, { "grad_norm": 0.35061466693878174, "learning_rate": 9.120568506400873e-05, "loss": 0.035, "step": 23210 }, { "grad_norm": 0.4746181070804596, "learning_rate": 9.119631715449548e-05, "loss": 0.0398, "step": 23220 }, { "grad_norm": 0.5864694118499756, "learning_rate": 9.118694473981493e-05, "loss": 0.0363, "step": 23230 }, { "grad_norm": 0.29706451296806335, "learning_rate": 9.117756782099203e-05, "loss": 0.0422, "step": 23240 }, { "grad_norm": 0.2989655137062073, "learning_rate": 9.11681863990522e-05, "loss": 0.0317, "step": 23250 }, { "grad_norm": 0.3207419812679291, "learning_rate": 9.115880047502142e-05, "loss": 0.0282, "step": 23260 }, { "grad_norm": 0.28389886021614075, "learning_rate": 9.114941004992609e-05, "loss": 0.0364, "step": 23270 }, { "grad_norm": 0.32389163970947266, "learning_rate": 9.114001512479317e-05, "loss": 0.0434, "step": 23280 }, { "grad_norm": 0.30144205689430237, "learning_rate": 9.113061570065003e-05, "loss": 0.0419, "step": 23290 }, { "grad_norm": 0.33398064970970154, "learning_rate": 9.112121177852459e-05, "loss": 0.0398, "step": 23300 }, { "grad_norm": 0.5528998374938965, "learning_rate": 9.111180335944527e-05, "loss": 0.0482, "step": 23310 }, { "grad_norm": 0.27592748403549194, "learning_rate": 9.110239044444093e-05, "loss": 0.0319, "step": 23320 }, { "grad_norm": 0.21351774036884308, "learning_rate": 9.109297303454099e-05, "loss": 0.026, "step": 23330 }, { "grad_norm": 0.41914182901382446, "learning_rate": 9.108355113077526e-05, "loss": 0.044, "step": 23340 }, { "grad_norm": 0.2355162352323532, "learning_rate": 9.107412473417419e-05, "loss": 0.0305, "step": 23350 }, { "grad_norm": 0.40856626629829407, "learning_rate": 9.106469384576858e-05, "loss": 0.0321, "step": 23360 }, { "grad_norm": 0.26126837730407715, "learning_rate": 9.105525846658978e-05, "loss": 0.031, "step": 23370 }, { "grad_norm": 0.3330550491809845, "learning_rate": 9.104581859766965e-05, "loss": 0.0352, "step": 23380 }, { "grad_norm": 0.33896154165267944, "learning_rate": 9.10363742400405e-05, "loss": 0.0434, "step": 23390 }, { "grad_norm": 0.3001539707183838, "learning_rate": 9.102692539473518e-05, "loss": 0.0445, "step": 23400 }, { "grad_norm": 0.37534329295158386, "learning_rate": 9.101747206278697e-05, "loss": 0.0491, "step": 23410 }, { "grad_norm": 0.277829110622406, "learning_rate": 9.100801424522968e-05, "loss": 0.0291, "step": 23420 }, { "grad_norm": 0.3458576500415802, "learning_rate": 9.099855194309762e-05, "loss": 0.0342, "step": 23430 }, { "grad_norm": 0.4076205790042877, "learning_rate": 9.098908515742554e-05, "loss": 0.0303, "step": 23440 }, { "grad_norm": 0.2198483645915985, "learning_rate": 9.097961388924873e-05, "loss": 0.0287, "step": 23450 }, { "grad_norm": 0.27767106890678406, "learning_rate": 9.097013813960298e-05, "loss": 0.0415, "step": 23460 }, { "grad_norm": 0.3687403202056885, "learning_rate": 9.09606579095245e-05, "loss": 0.0269, "step": 23470 }, { "grad_norm": 0.22635425627231598, "learning_rate": 9.095117320005008e-05, "loss": 0.0352, "step": 23480 }, { "grad_norm": 0.2532857060432434, "learning_rate": 9.094168401221691e-05, "loss": 0.0331, "step": 23490 }, { "grad_norm": 0.32971516251564026, "learning_rate": 9.093219034706273e-05, "loss": 0.0488, "step": 23500 }, { "grad_norm": 0.25137996673583984, "learning_rate": 9.092269220562577e-05, "loss": 0.0392, "step": 23510 }, { "grad_norm": 0.35645630955696106, "learning_rate": 9.09131895889447e-05, "loss": 0.0521, "step": 23520 }, { "grad_norm": 0.33362269401550293, "learning_rate": 9.090368249805873e-05, "loss": 0.0341, "step": 23530 }, { "grad_norm": 0.2645956873893738, "learning_rate": 9.089417093400754e-05, "loss": 0.0281, "step": 23540 }, { "grad_norm": 0.3712826371192932, "learning_rate": 9.088465489783131e-05, "loss": 0.0272, "step": 23550 }, { "grad_norm": 0.23418544232845306, "learning_rate": 9.087513439057068e-05, "loss": 0.0216, "step": 23560 }, { "grad_norm": 0.4555492699146271, "learning_rate": 9.08656094132668e-05, "loss": 0.0376, "step": 23570 }, { "grad_norm": 0.3604148328304291, "learning_rate": 9.085607996696134e-05, "loss": 0.0326, "step": 23580 }, { "grad_norm": 0.2859329581260681, "learning_rate": 9.084654605269639e-05, "loss": 0.03, "step": 23590 }, { "grad_norm": 0.40391209721565247, "learning_rate": 9.083700767151457e-05, "loss": 0.0287, "step": 23600 }, { "grad_norm": 0.7991213798522949, "learning_rate": 9.082746482445898e-05, "loss": 0.0337, "step": 23610 }, { "grad_norm": 0.25544822216033936, "learning_rate": 9.081791751257325e-05, "loss": 0.0282, "step": 23620 }, { "grad_norm": 0.3595968186855316, "learning_rate": 9.080836573690142e-05, "loss": 0.0416, "step": 23630 }, { "grad_norm": 0.22096781432628632, "learning_rate": 9.079880949848805e-05, "loss": 0.0444, "step": 23640 }, { "grad_norm": 0.19745506346225739, "learning_rate": 9.078924879837822e-05, "loss": 0.0364, "step": 23650 }, { "grad_norm": 0.31220871210098267, "learning_rate": 9.077968363761747e-05, "loss": 0.0336, "step": 23660 }, { "grad_norm": 0.22607854008674622, "learning_rate": 9.077011401725182e-05, "loss": 0.0249, "step": 23670 }, { "grad_norm": 0.27535930275917053, "learning_rate": 9.07605399383278e-05, "loss": 0.0363, "step": 23680 }, { "grad_norm": 0.24695035815238953, "learning_rate": 9.075096140189243e-05, "loss": 0.03, "step": 23690 }, { "grad_norm": 0.6298685073852539, "learning_rate": 9.074137840899318e-05, "loss": 0.0439, "step": 23700 }, { "grad_norm": 0.42842355370521545, "learning_rate": 9.073179096067804e-05, "loss": 0.0309, "step": 23710 }, { "grad_norm": 0.3809994161128998, "learning_rate": 9.072219905799549e-05, "loss": 0.0373, "step": 23720 }, { "grad_norm": 0.506632924079895, "learning_rate": 9.071260270199447e-05, "loss": 0.0461, "step": 23730 }, { "grad_norm": 0.40398553013801575, "learning_rate": 9.070300189372441e-05, "loss": 0.0296, "step": 23740 }, { "grad_norm": 0.23592129349708557, "learning_rate": 9.069339663423528e-05, "loss": 0.0391, "step": 23750 }, { "grad_norm": 0.3393288850784302, "learning_rate": 9.068378692457747e-05, "loss": 0.0467, "step": 23760 }, { "grad_norm": 0.352963387966156, "learning_rate": 9.067417276580189e-05, "loss": 0.0381, "step": 23770 }, { "grad_norm": 0.26497185230255127, "learning_rate": 9.066455415895993e-05, "loss": 0.031, "step": 23780 }, { "grad_norm": 0.4156630337238312, "learning_rate": 9.065493110510346e-05, "loss": 0.0367, "step": 23790 }, { "grad_norm": 0.20514731109142303, "learning_rate": 9.064530360528484e-05, "loss": 0.0307, "step": 23800 }, { "grad_norm": 0.3794015049934387, "learning_rate": 9.063567166055695e-05, "loss": 0.0388, "step": 23810 }, { "grad_norm": 0.2820889949798584, "learning_rate": 9.062603527197308e-05, "loss": 0.031, "step": 23820 }, { "grad_norm": 0.337272047996521, "learning_rate": 9.06163944405871e-05, "loss": 0.0308, "step": 23830 }, { "grad_norm": 0.37890613079071045, "learning_rate": 9.060674916745327e-05, "loss": 0.0317, "step": 23840 }, { "grad_norm": 0.43008410930633545, "learning_rate": 9.05970994536264e-05, "loss": 0.0216, "step": 23850 }, { "grad_norm": 0.23499183356761932, "learning_rate": 9.05874453001618e-05, "loss": 0.0271, "step": 23860 }, { "grad_norm": 0.2507721185684204, "learning_rate": 9.057778670811517e-05, "loss": 0.0265, "step": 23870 }, { "grad_norm": 0.390388160943985, "learning_rate": 9.056812367854281e-05, "loss": 0.0276, "step": 23880 }, { "grad_norm": 0.17574460804462433, "learning_rate": 9.055845621250143e-05, "loss": 0.032, "step": 23890 }, { "grad_norm": 0.1824025809764862, "learning_rate": 9.054878431104825e-05, "loss": 0.0228, "step": 23900 }, { "grad_norm": 0.35238879919052124, "learning_rate": 9.0539107975241e-05, "loss": 0.0285, "step": 23910 }, { "grad_norm": 0.4145505428314209, "learning_rate": 9.052942720613784e-05, "loss": 0.045, "step": 23920 }, { "grad_norm": 0.3093509376049042, "learning_rate": 9.051974200479745e-05, "loss": 0.0279, "step": 23930 }, { "grad_norm": 0.30891308188438416, "learning_rate": 9.051005237227901e-05, "loss": 0.0225, "step": 23940 }, { "grad_norm": 0.16766177117824554, "learning_rate": 9.050035830964215e-05, "loss": 0.0233, "step": 23950 }, { "grad_norm": 0.33055731654167175, "learning_rate": 9.049065981794698e-05, "loss": 0.0384, "step": 23960 }, { "grad_norm": 0.2713353931903839, "learning_rate": 9.048095689825414e-05, "loss": 0.0317, "step": 23970 }, { "grad_norm": 0.3764590322971344, "learning_rate": 9.047124955162472e-05, "loss": 0.0352, "step": 23980 }, { "grad_norm": 0.27533024549484253, "learning_rate": 9.046153777912028e-05, "loss": 0.0422, "step": 23990 }, { "grad_norm": 0.33963146805763245, "learning_rate": 9.045182158180292e-05, "loss": 0.0269, "step": 24000 }, { "grad_norm": 0.23715810477733612, "learning_rate": 9.044210096073516e-05, "loss": 0.0224, "step": 24010 }, { "grad_norm": 0.2172412872314453, "learning_rate": 9.043237591698004e-05, "loss": 0.0301, "step": 24020 }, { "grad_norm": 0.2759350538253784, "learning_rate": 9.04226464516011e-05, "loss": 0.0331, "step": 24030 }, { "grad_norm": 0.3173430860042572, "learning_rate": 9.041291256566229e-05, "loss": 0.034, "step": 24040 }, { "grad_norm": 0.230209082365036, "learning_rate": 9.040317426022814e-05, "loss": 0.032, "step": 24050 }, { "grad_norm": 0.3750130534172058, "learning_rate": 9.03934315363636e-05, "loss": 0.0344, "step": 24060 }, { "grad_norm": 0.16679294407367706, "learning_rate": 9.038368439513409e-05, "loss": 0.0258, "step": 24070 }, { "grad_norm": 0.19318264722824097, "learning_rate": 9.03739328376056e-05, "loss": 0.0345, "step": 24080 }, { "grad_norm": 0.22969499230384827, "learning_rate": 9.036417686484451e-05, "loss": 0.0246, "step": 24090 }, { "grad_norm": 0.4938340485095978, "learning_rate": 9.035441647791773e-05, "loss": 0.0317, "step": 24100 }, { "grad_norm": 0.1478634774684906, "learning_rate": 9.034465167789263e-05, "loss": 0.0247, "step": 24110 }, { "grad_norm": 0.2625941038131714, "learning_rate": 9.033488246583706e-05, "loss": 0.0279, "step": 24120 }, { "grad_norm": 0.18844804167747498, "learning_rate": 9.032510884281941e-05, "loss": 0.031, "step": 24130 }, { "grad_norm": 0.19735439121723175, "learning_rate": 9.031533080990848e-05, "loss": 0.0401, "step": 24140 }, { "grad_norm": 0.4380514323711395, "learning_rate": 9.030554836817358e-05, "loss": 0.0331, "step": 24150 }, { "grad_norm": 0.33156725764274597, "learning_rate": 9.029576151868451e-05, "loss": 0.0359, "step": 24160 }, { "grad_norm": 0.1765851527452469, "learning_rate": 9.028597026251155e-05, "loss": 0.0354, "step": 24170 }, { "grad_norm": 0.28044626116752625, "learning_rate": 9.027617460072547e-05, "loss": 0.0261, "step": 24180 }, { "grad_norm": 0.29983630776405334, "learning_rate": 9.026637453439745e-05, "loss": 0.0314, "step": 24190 }, { "grad_norm": 0.35047221183776855, "learning_rate": 9.025657006459927e-05, "loss": 0.0213, "step": 24200 }, { "grad_norm": 0.39230722188949585, "learning_rate": 9.024676119240311e-05, "loss": 0.026, "step": 24210 }, { "grad_norm": 0.21546117961406708, "learning_rate": 9.023694791888166e-05, "loss": 0.023, "step": 24220 }, { "grad_norm": 0.33584538102149963, "learning_rate": 9.022713024510808e-05, "loss": 0.0298, "step": 24230 }, { "grad_norm": 0.36306431889533997, "learning_rate": 9.021730817215601e-05, "loss": 0.0281, "step": 24240 }, { "grad_norm": 0.3374752998352051, "learning_rate": 9.02074817010996e-05, "loss": 0.031, "step": 24250 }, { "grad_norm": 0.4293327033519745, "learning_rate": 9.019765083301342e-05, "loss": 0.0375, "step": 24260 }, { "grad_norm": 0.268228143453598, "learning_rate": 9.01878155689726e-05, "loss": 0.0226, "step": 24270 }, { "grad_norm": 0.4676518440246582, "learning_rate": 9.017797591005268e-05, "loss": 0.0386, "step": 24280 }, { "grad_norm": 0.2952321469783783, "learning_rate": 9.016813185732972e-05, "loss": 0.0365, "step": 24290 }, { "grad_norm": 0.3382284939289093, "learning_rate": 9.015828341188027e-05, "loss": 0.0323, "step": 24300 }, { "grad_norm": 0.3059752583503723, "learning_rate": 9.01484305747813e-05, "loss": 0.0322, "step": 24310 }, { "grad_norm": 0.28237321972846985, "learning_rate": 9.013857334711033e-05, "loss": 0.0241, "step": 24320 }, { "grad_norm": 0.36669492721557617, "learning_rate": 9.012871172994534e-05, "loss": 0.0269, "step": 24330 }, { "grad_norm": 0.2854349613189697, "learning_rate": 9.011884572436476e-05, "loss": 0.0318, "step": 24340 }, { "grad_norm": 0.3425736427307129, "learning_rate": 9.010897533144754e-05, "loss": 0.0375, "step": 24350 }, { "grad_norm": 0.21279005706310272, "learning_rate": 9.009910055227306e-05, "loss": 0.0279, "step": 24360 }, { "grad_norm": 0.16272324323654175, "learning_rate": 9.008922138792124e-05, "loss": 0.0317, "step": 24370 }, { "grad_norm": 0.210994154214859, "learning_rate": 9.007933783947244e-05, "loss": 0.0224, "step": 24380 }, { "grad_norm": 0.5602737069129944, "learning_rate": 9.006944990800752e-05, "loss": 0.029, "step": 24390 }, { "grad_norm": 0.15274275839328766, "learning_rate": 9.005955759460779e-05, "loss": 0.03, "step": 24400 }, { "grad_norm": 0.3728621304035187, "learning_rate": 9.004966090035508e-05, "loss": 0.0298, "step": 24410 }, { "grad_norm": 0.4870375096797943, "learning_rate": 9.003975982633166e-05, "loss": 0.0512, "step": 24420 }, { "grad_norm": 0.28786128759384155, "learning_rate": 9.00298543736203e-05, "loss": 0.0459, "step": 24430 }, { "grad_norm": 0.24030719697475433, "learning_rate": 9.001994454330427e-05, "loss": 0.0276, "step": 24440 }, { "grad_norm": 0.4501763582229614, "learning_rate": 9.001003033646727e-05, "loss": 0.0376, "step": 24450 }, { "grad_norm": 0.38935333490371704, "learning_rate": 9.00001117541935e-05, "loss": 0.0334, "step": 24460 }, { "grad_norm": 0.32979539036750793, "learning_rate": 8.999018879756764e-05, "loss": 0.0324, "step": 24470 }, { "grad_norm": 0.5654341578483582, "learning_rate": 8.998026146767487e-05, "loss": 0.0411, "step": 24480 }, { "grad_norm": 0.30614596605300903, "learning_rate": 8.99703297656008e-05, "loss": 0.0282, "step": 24490 }, { "grad_norm": 0.36623796820640564, "learning_rate": 8.996039369243156e-05, "loss": 0.0409, "step": 24500 }, { "grad_norm": 0.2719048857688904, "learning_rate": 8.995045324925378e-05, "loss": 0.0292, "step": 24510 }, { "grad_norm": 0.2387358397245407, "learning_rate": 8.994050843715448e-05, "loss": 0.0259, "step": 24520 }, { "grad_norm": 0.23087440431118011, "learning_rate": 8.993055925722121e-05, "loss": 0.0278, "step": 24530 }, { "grad_norm": 0.2668653428554535, "learning_rate": 8.992060571054202e-05, "loss": 0.0227, "step": 24540 }, { "grad_norm": 0.24165889620780945, "learning_rate": 8.991064779820542e-05, "loss": 0.037, "step": 24550 }, { "grad_norm": 0.245470330119133, "learning_rate": 8.990068552130036e-05, "loss": 0.0254, "step": 24560 }, { "grad_norm": 0.19424527883529663, "learning_rate": 8.989071888091634e-05, "loss": 0.0204, "step": 24570 }, { "grad_norm": 0.22620819509029388, "learning_rate": 8.988074787814329e-05, "loss": 0.0186, "step": 24580 }, { "grad_norm": 0.25248533487319946, "learning_rate": 8.987077251407158e-05, "loss": 0.0246, "step": 24590 }, { "grad_norm": 0.15701298415660858, "learning_rate": 8.986079278979216e-05, "loss": 0.0195, "step": 24600 }, { "grad_norm": 0.36038753390312195, "learning_rate": 8.985080870639635e-05, "loss": 0.0297, "step": 24610 }, { "grad_norm": 0.3368123769760132, "learning_rate": 8.984082026497603e-05, "loss": 0.0285, "step": 24620 }, { "grad_norm": 0.3866752088069916, "learning_rate": 8.98308274666235e-05, "loss": 0.0345, "step": 24630 }, { "grad_norm": 0.3482781648635864, "learning_rate": 8.982083031243155e-05, "loss": 0.0277, "step": 24640 }, { "grad_norm": 0.28684157133102417, "learning_rate": 8.98108288034935e-05, "loss": 0.0324, "step": 24650 }, { "grad_norm": 0.2852974236011505, "learning_rate": 8.980082294090305e-05, "loss": 0.0389, "step": 24660 }, { "grad_norm": 0.26335570216178894, "learning_rate": 8.979081272575443e-05, "loss": 0.0351, "step": 24670 }, { "grad_norm": 0.49344533681869507, "learning_rate": 8.978079815914236e-05, "loss": 0.0296, "step": 24680 }, { "grad_norm": 0.27999168634414673, "learning_rate": 8.977077924216202e-05, "loss": 0.0447, "step": 24690 }, { "grad_norm": 0.5224990248680115, "learning_rate": 8.976075597590905e-05, "loss": 0.0402, "step": 24700 }, { "grad_norm": 0.5170624852180481, "learning_rate": 8.975072836147958e-05, "loss": 0.0296, "step": 24710 }, { "grad_norm": 0.2961171567440033, "learning_rate": 8.974069639997025e-05, "loss": 0.0383, "step": 24720 }, { "grad_norm": 0.6131241917610168, "learning_rate": 8.973066009247808e-05, "loss": 0.038, "step": 24730 }, { "grad_norm": 0.2899106442928314, "learning_rate": 8.972061944010066e-05, "loss": 0.0368, "step": 24740 }, { "grad_norm": 0.5705959796905518, "learning_rate": 8.971057444393603e-05, "loss": 0.0369, "step": 24750 }, { "grad_norm": 0.3611326217651367, "learning_rate": 8.970052510508268e-05, "loss": 0.0338, "step": 24760 }, { "grad_norm": 0.23672722280025482, "learning_rate": 8.969047142463959e-05, "loss": 0.0365, "step": 24770 }, { "grad_norm": 0.6194960474967957, "learning_rate": 8.968041340370621e-05, "loss": 0.038, "step": 24780 }, { "grad_norm": 0.2860132157802582, "learning_rate": 8.96703510433825e-05, "loss": 0.0309, "step": 24790 }, { "grad_norm": 0.21818988025188446, "learning_rate": 8.966028434476883e-05, "loss": 0.026, "step": 24800 }, { "grad_norm": 0.44149962067604065, "learning_rate": 8.96502133089661e-05, "loss": 0.0392, "step": 24810 }, { "grad_norm": 0.5596978664398193, "learning_rate": 8.964013793707564e-05, "loss": 0.0335, "step": 24820 }, { "grad_norm": 0.12405805289745331, "learning_rate": 8.963005823019932e-05, "loss": 0.0251, "step": 24830 }, { "grad_norm": 0.5228030681610107, "learning_rate": 8.961997418943939e-05, "loss": 0.0331, "step": 24840 }, { "grad_norm": 0.4099101424217224, "learning_rate": 8.960988581589865e-05, "loss": 0.0339, "step": 24850 }, { "grad_norm": 0.2939956486225128, "learning_rate": 8.959979311068037e-05, "loss": 0.0299, "step": 24860 }, { "grad_norm": 0.22844161093235016, "learning_rate": 8.958969607488823e-05, "loss": 0.0318, "step": 24870 }, { "grad_norm": 0.322052001953125, "learning_rate": 8.957959470962647e-05, "loss": 0.0291, "step": 24880 }, { "grad_norm": 0.9411081075668335, "learning_rate": 8.956948901599971e-05, "loss": 0.0385, "step": 24890 }, { "grad_norm": 0.4594387710094452, "learning_rate": 8.955937899511315e-05, "loss": 0.0409, "step": 24900 }, { "grad_norm": 0.38833507895469666, "learning_rate": 8.954926464807238e-05, "loss": 0.0385, "step": 24910 }, { "grad_norm": 0.22998318076133728, "learning_rate": 8.953914597598347e-05, "loss": 0.0316, "step": 24920 }, { "grad_norm": 0.210226371884346, "learning_rate": 8.952902297995303e-05, "loss": 0.032, "step": 24930 }, { "grad_norm": 0.2478412687778473, "learning_rate": 8.951889566108804e-05, "loss": 0.0402, "step": 24940 }, { "grad_norm": 0.16304844617843628, "learning_rate": 8.950876402049606e-05, "loss": 0.043, "step": 24950 }, { "grad_norm": 0.31280267238616943, "learning_rate": 8.949862805928504e-05, "loss": 0.0293, "step": 24960 }, { "grad_norm": 0.14953601360321045, "learning_rate": 8.948848777856343e-05, "loss": 0.0298, "step": 24970 }, { "grad_norm": 0.377765953540802, "learning_rate": 8.947834317944017e-05, "loss": 0.0286, "step": 24980 }, { "grad_norm": 0.5137674808502197, "learning_rate": 8.946819426302466e-05, "loss": 0.0397, "step": 24990 }, { "grad_norm": 0.7398988008499146, "learning_rate": 8.945804103042676e-05, "loss": 0.0316, "step": 25000 }, { "grad_norm": 0.1945265531539917, "learning_rate": 8.944788348275681e-05, "loss": 0.025, "step": 25010 }, { "grad_norm": 0.21814703941345215, "learning_rate": 8.943772162112565e-05, "loss": 0.036, "step": 25020 }, { "grad_norm": 0.32587680220603943, "learning_rate": 8.942755544664454e-05, "loss": 0.0379, "step": 25030 }, { "grad_norm": 0.26388469338417053, "learning_rate": 8.941738496042525e-05, "loss": 0.0194, "step": 25040 }, { "grad_norm": 0.306701123714447, "learning_rate": 8.940721016357999e-05, "loss": 0.0342, "step": 25050 }, { "grad_norm": 0.2765086591243744, "learning_rate": 8.939703105722148e-05, "loss": 0.0305, "step": 25060 }, { "grad_norm": 0.2532218396663666, "learning_rate": 8.93868476424629e-05, "loss": 0.0237, "step": 25070 }, { "grad_norm": 0.2307414710521698, "learning_rate": 8.937665992041786e-05, "loss": 0.0278, "step": 25080 }, { "grad_norm": 0.4162510633468628, "learning_rate": 8.93664678922005e-05, "loss": 0.0399, "step": 25090 }, { "grad_norm": 0.42073848843574524, "learning_rate": 8.93562715589254e-05, "loss": 0.0368, "step": 25100 }, { "grad_norm": 0.32374146580696106, "learning_rate": 8.934607092170762e-05, "loss": 0.0245, "step": 25110 }, { "grad_norm": 0.3353014886379242, "learning_rate": 8.933586598166266e-05, "loss": 0.0277, "step": 25120 }, { "grad_norm": 0.18873801827430725, "learning_rate": 8.932565673990655e-05, "loss": 0.0472, "step": 25130 }, { "grad_norm": 0.23661372065544128, "learning_rate": 8.931544319755574e-05, "loss": 0.0285, "step": 25140 }, { "grad_norm": 0.23598799109458923, "learning_rate": 8.930522535572718e-05, "loss": 0.0286, "step": 25150 }, { "grad_norm": 0.2792911231517792, "learning_rate": 8.929500321553826e-05, "loss": 0.0293, "step": 25160 }, { "grad_norm": 0.2775035500526428, "learning_rate": 8.928477677810686e-05, "loss": 0.0256, "step": 25170 }, { "grad_norm": 0.4463295340538025, "learning_rate": 8.927454604455137e-05, "loss": 0.0383, "step": 25180 }, { "grad_norm": 0.2808997631072998, "learning_rate": 8.926431101599053e-05, "loss": 0.0371, "step": 25190 }, { "grad_norm": 0.18430332839488983, "learning_rate": 8.925407169354369e-05, "loss": 0.0272, "step": 25200 }, { "grad_norm": 0.2836666703224182, "learning_rate": 8.92438280783306e-05, "loss": 0.0352, "step": 25210 }, { "grad_norm": 0.28719469904899597, "learning_rate": 8.923358017147146e-05, "loss": 0.0306, "step": 25220 }, { "grad_norm": 0.5561537742614746, "learning_rate": 8.922332797408697e-05, "loss": 0.0442, "step": 25230 }, { "grad_norm": 0.2558865547180176, "learning_rate": 8.921307148729831e-05, "loss": 0.0327, "step": 25240 }, { "grad_norm": 0.2533521354198456, "learning_rate": 8.920281071222712e-05, "loss": 0.0274, "step": 25250 }, { "grad_norm": 0.42252928018569946, "learning_rate": 8.919254564999548e-05, "loss": 0.031, "step": 25260 }, { "grad_norm": 0.35801491141319275, "learning_rate": 8.918227630172598e-05, "loss": 0.0347, "step": 25270 }, { "grad_norm": 0.2109207659959793, "learning_rate": 8.917200266854165e-05, "loss": 0.036, "step": 25280 }, { "grad_norm": 0.7817487716674805, "learning_rate": 8.9161724751566e-05, "loss": 0.0296, "step": 25290 }, { "grad_norm": 0.42529067397117615, "learning_rate": 8.915144255192302e-05, "loss": 0.0376, "step": 25300 }, { "grad_norm": 0.633806049823761, "learning_rate": 8.914115607073714e-05, "loss": 0.035, "step": 25310 }, { "grad_norm": 0.45417317748069763, "learning_rate": 8.913086530913327e-05, "loss": 0.0232, "step": 25320 }, { "grad_norm": 0.298749715089798, "learning_rate": 8.912057026823681e-05, "loss": 0.0526, "step": 25330 }, { "grad_norm": 0.14667752385139465, "learning_rate": 8.91102709491736e-05, "loss": 0.0218, "step": 25340 }, { "grad_norm": 0.23697002232074738, "learning_rate": 8.909996735306996e-05, "loss": 0.0309, "step": 25350 }, { "grad_norm": 0.22255194187164307, "learning_rate": 8.908965948105268e-05, "loss": 0.0336, "step": 25360 }, { "grad_norm": 0.31952860951423645, "learning_rate": 8.907934733424901e-05, "loss": 0.0359, "step": 25370 }, { "grad_norm": 0.19232460856437683, "learning_rate": 8.906903091378666e-05, "loss": 0.0269, "step": 25380 }, { "grad_norm": 0.2132520228624344, "learning_rate": 8.905871022079384e-05, "loss": 0.0248, "step": 25390 }, { "grad_norm": 0.2949303686618805, "learning_rate": 8.90483852563992e-05, "loss": 0.0319, "step": 25400 }, { "grad_norm": 0.27180880308151245, "learning_rate": 8.903805602173185e-05, "loss": 0.0299, "step": 25410 }, { "grad_norm": 0.35547876358032227, "learning_rate": 8.902772251792137e-05, "loss": 0.0308, "step": 25420 }, { "grad_norm": 0.4365587532520294, "learning_rate": 8.901738474609786e-05, "loss": 0.0333, "step": 25430 }, { "grad_norm": 0.8297638893127441, "learning_rate": 8.900704270739179e-05, "loss": 0.0458, "step": 25440 }, { "grad_norm": 0.40575122833251953, "learning_rate": 8.89966964029342e-05, "loss": 0.0303, "step": 25450 }, { "grad_norm": 0.2839295268058777, "learning_rate": 8.898634583385652e-05, "loss": 0.0276, "step": 25460 }, { "grad_norm": 0.3205222189426422, "learning_rate": 8.897599100129065e-05, "loss": 0.0304, "step": 25470 }, { "grad_norm": 0.19939416646957397, "learning_rate": 8.896563190636903e-05, "loss": 0.029, "step": 25480 }, { "grad_norm": 0.4422474205493927, "learning_rate": 8.895526855022448e-05, "loss": 0.0276, "step": 25490 }, { "grad_norm": 0.2592502236366272, "learning_rate": 8.894490093399033e-05, "loss": 0.0384, "step": 25500 }, { "grad_norm": 0.1444052904844284, "learning_rate": 8.893452905880035e-05, "loss": 0.0267, "step": 25510 }, { "grad_norm": 0.35121870040893555, "learning_rate": 8.892415292578883e-05, "loss": 0.0409, "step": 25520 }, { "grad_norm": 0.2821705937385559, "learning_rate": 8.891377253609046e-05, "loss": 0.0379, "step": 25530 }, { "grad_norm": 0.49878838658332825, "learning_rate": 8.890338789084043e-05, "loss": 0.026, "step": 25540 }, { "grad_norm": 0.32517722249031067, "learning_rate": 8.88929989911744e-05, "loss": 0.0282, "step": 25550 }, { "grad_norm": 0.2783792018890381, "learning_rate": 8.888260583822847e-05, "loss": 0.0433, "step": 25560 }, { "grad_norm": 0.3295013904571533, "learning_rate": 8.887220843313921e-05, "loss": 0.0314, "step": 25570 }, { "grad_norm": 0.25965240597724915, "learning_rate": 8.88618067770437e-05, "loss": 0.0317, "step": 25580 }, { "grad_norm": 0.2754440903663635, "learning_rate": 8.885140087107942e-05, "loss": 0.0312, "step": 25590 }, { "grad_norm": 0.5123128890991211, "learning_rate": 8.884099071638436e-05, "loss": 0.0385, "step": 25600 }, { "grad_norm": 0.20682765543460846, "learning_rate": 8.883057631409695e-05, "loss": 0.0227, "step": 25610 }, { "grad_norm": 0.22212989628314972, "learning_rate": 8.882015766535608e-05, "loss": 0.0337, "step": 25620 }, { "grad_norm": 0.4026961326599121, "learning_rate": 8.880973477130115e-05, "loss": 0.0313, "step": 25630 }, { "grad_norm": 0.1737818717956543, "learning_rate": 8.879930763307197e-05, "loss": 0.0168, "step": 25640 }, { "grad_norm": 0.2260996699333191, "learning_rate": 8.878887625180884e-05, "loss": 0.0298, "step": 25650 }, { "grad_norm": 0.19313788414001465, "learning_rate": 8.877844062865253e-05, "loss": 0.0163, "step": 25660 }, { "grad_norm": 0.39824235439300537, "learning_rate": 8.876800076474424e-05, "loss": 0.0342, "step": 25670 }, { "grad_norm": 0.23548460006713867, "learning_rate": 8.875755666122568e-05, "loss": 0.0289, "step": 25680 }, { "grad_norm": 0.24441322684288025, "learning_rate": 8.8747108319239e-05, "loss": 0.0316, "step": 25690 }, { "grad_norm": 0.3504754304885864, "learning_rate": 8.87366557399268e-05, "loss": 0.0297, "step": 25700 }, { "grad_norm": 0.2966170907020569, "learning_rate": 8.872619892443217e-05, "loss": 0.0292, "step": 25710 }, { "grad_norm": 0.3188092112541199, "learning_rate": 8.871573787389865e-05, "loss": 0.0226, "step": 25720 }, { "grad_norm": 0.19399575889110565, "learning_rate": 8.870527258947024e-05, "loss": 0.023, "step": 25730 }, { "grad_norm": 0.18788209557533264, "learning_rate": 8.869480307229143e-05, "loss": 0.0241, "step": 25740 }, { "grad_norm": 0.21853309869766235, "learning_rate": 8.868432932350712e-05, "loss": 0.0223, "step": 25750 }, { "grad_norm": 0.42619916796684265, "learning_rate": 8.867385134426272e-05, "loss": 0.029, "step": 25760 }, { "grad_norm": 0.14018569886684418, "learning_rate": 8.866336913570407e-05, "loss": 0.0198, "step": 25770 }, { "grad_norm": 0.34265387058258057, "learning_rate": 8.865288269897751e-05, "loss": 0.0291, "step": 25780 }, { "grad_norm": 0.24567487835884094, "learning_rate": 8.864239203522981e-05, "loss": 0.0329, "step": 25790 }, { "grad_norm": 0.3256150484085083, "learning_rate": 8.863189714560822e-05, "loss": 0.039, "step": 25800 }, { "grad_norm": 0.32075488567352295, "learning_rate": 8.862139803126043e-05, "loss": 0.029, "step": 25810 }, { "grad_norm": 0.35602736473083496, "learning_rate": 8.861089469333463e-05, "loss": 0.0365, "step": 25820 }, { "grad_norm": 0.19481566548347473, "learning_rate": 8.860038713297944e-05, "loss": 0.0234, "step": 25830 }, { "grad_norm": 0.27083542943000793, "learning_rate": 8.858987535134394e-05, "loss": 0.0345, "step": 25840 }, { "grad_norm": 0.6771640181541443, "learning_rate": 8.857935934957769e-05, "loss": 0.036, "step": 25850 }, { "grad_norm": 0.27383360266685486, "learning_rate": 8.856883912883071e-05, "loss": 0.0306, "step": 25860 }, { "grad_norm": 0.2002677470445633, "learning_rate": 8.855831469025346e-05, "loss": 0.0228, "step": 25870 }, { "grad_norm": 0.3105581998825073, "learning_rate": 8.854778603499689e-05, "loss": 0.0324, "step": 25880 }, { "grad_norm": 0.6312800645828247, "learning_rate": 8.85372531642124e-05, "loss": 0.0398, "step": 25890 }, { "grad_norm": 0.35552895069122314, "learning_rate": 8.852671607905185e-05, "loss": 0.03, "step": 25900 }, { "grad_norm": 0.24119549989700317, "learning_rate": 8.851617478066754e-05, "loss": 0.0284, "step": 25910 }, { "grad_norm": 0.4833205044269562, "learning_rate": 8.850562927021227e-05, "loss": 0.0323, "step": 25920 }, { "grad_norm": 0.34723928570747375, "learning_rate": 8.849507954883928e-05, "loss": 0.0226, "step": 25930 }, { "grad_norm": 0.276638925075531, "learning_rate": 8.848452561770226e-05, "loss": 0.0201, "step": 25940 }, { "grad_norm": 0.2501017153263092, "learning_rate": 8.847396747795538e-05, "loss": 0.0359, "step": 25950 }, { "grad_norm": 0.20773369073867798, "learning_rate": 8.846340513075327e-05, "loss": 0.0299, "step": 25960 }, { "grad_norm": 0.2890189588069916, "learning_rate": 8.845283857725099e-05, "loss": 0.024, "step": 25970 }, { "grad_norm": 0.23943497240543365, "learning_rate": 8.844226781860409e-05, "loss": 0.0237, "step": 25980 }, { "grad_norm": 0.1027279645204544, "learning_rate": 8.84316928559686e-05, "loss": 0.0201, "step": 25990 }, { "grad_norm": 0.45309868454933167, "learning_rate": 8.842111369050094e-05, "loss": 0.0267, "step": 26000 }, { "grad_norm": 0.3323332667350769, "learning_rate": 8.841053032335808e-05, "loss": 0.0255, "step": 26010 }, { "grad_norm": 0.281394898891449, "learning_rate": 8.839994275569735e-05, "loss": 0.0215, "step": 26020 }, { "grad_norm": 0.47378477454185486, "learning_rate": 8.838935098867662e-05, "loss": 0.0433, "step": 26030 }, { "grad_norm": 0.23828816413879395, "learning_rate": 8.837875502345418e-05, "loss": 0.035, "step": 26040 }, { "grad_norm": 0.22998832166194916, "learning_rate": 8.83681548611888e-05, "loss": 0.0341, "step": 26050 }, { "grad_norm": 0.22466403245925903, "learning_rate": 8.835755050303969e-05, "loss": 0.0196, "step": 26060 }, { "grad_norm": 0.25983384251594543, "learning_rate": 8.834694195016653e-05, "loss": 0.0256, "step": 26070 }, { "grad_norm": 0.2978363037109375, "learning_rate": 8.833632920372942e-05, "loss": 0.0374, "step": 26080 }, { "grad_norm": 0.3381209671497345, "learning_rate": 8.832571226488903e-05, "loss": 0.0295, "step": 26090 }, { "grad_norm": 0.37730735540390015, "learning_rate": 8.831509113480634e-05, "loss": 0.0318, "step": 26100 }, { "grad_norm": 0.21432891488075256, "learning_rate": 8.83044658146429e-05, "loss": 0.0276, "step": 26110 }, { "grad_norm": 0.2580643892288208, "learning_rate": 8.829383630556067e-05, "loss": 0.0286, "step": 26120 }, { "grad_norm": 0.385050892829895, "learning_rate": 8.828320260872207e-05, "loss": 0.0291, "step": 26130 }, { "grad_norm": 0.5225986242294312, "learning_rate": 8.827256472529e-05, "loss": 0.0366, "step": 26140 }, { "grad_norm": 0.3703421950340271, "learning_rate": 8.826192265642778e-05, "loss": 0.0324, "step": 26150 }, { "grad_norm": 0.2515425384044647, "learning_rate": 8.825127640329923e-05, "loss": 0.0333, "step": 26160 }, { "grad_norm": 0.2248619943857193, "learning_rate": 8.824062596706861e-05, "loss": 0.0335, "step": 26170 }, { "grad_norm": 0.3571884334087372, "learning_rate": 8.822997134890062e-05, "loss": 0.0364, "step": 26180 }, { "grad_norm": 0.454761266708374, "learning_rate": 8.821931254996044e-05, "loss": 0.0311, "step": 26190 }, { "grad_norm": 0.4937715232372284, "learning_rate": 8.82086495714137e-05, "loss": 0.0283, "step": 26200 }, { "grad_norm": 0.3268361985683441, "learning_rate": 8.81979824144265e-05, "loss": 0.0278, "step": 26210 }, { "grad_norm": 0.4349852204322815, "learning_rate": 8.818731108016536e-05, "loss": 0.0367, "step": 26220 }, { "grad_norm": 0.6117531657218933, "learning_rate": 8.81766355697973e-05, "loss": 0.0339, "step": 26230 }, { "grad_norm": 0.30886733531951904, "learning_rate": 8.816595588448977e-05, "loss": 0.0397, "step": 26240 }, { "grad_norm": 0.28840070962905884, "learning_rate": 8.81552720254107e-05, "loss": 0.027, "step": 26250 }, { "grad_norm": 0.46790656447410583, "learning_rate": 8.814458399372842e-05, "loss": 0.0373, "step": 26260 }, { "grad_norm": 0.20997528731822968, "learning_rate": 8.813389179061181e-05, "loss": 0.0157, "step": 26270 }, { "grad_norm": 0.5254262089729309, "learning_rate": 8.812319541723012e-05, "loss": 0.0434, "step": 26280 }, { "grad_norm": 0.2878764569759369, "learning_rate": 8.811249487475309e-05, "loss": 0.0517, "step": 26290 }, { "grad_norm": 0.2173728346824646, "learning_rate": 8.810179016435092e-05, "loss": 0.036, "step": 26300 }, { "grad_norm": 0.34617799520492554, "learning_rate": 8.809108128719428e-05, "loss": 0.0531, "step": 26310 }, { "grad_norm": 0.2125585824251175, "learning_rate": 8.808036824445424e-05, "loss": 0.0282, "step": 26320 }, { "grad_norm": 0.4872646927833557, "learning_rate": 8.806965103730238e-05, "loss": 0.0352, "step": 26330 }, { "grad_norm": 0.3572070896625519, "learning_rate": 8.805892966691074e-05, "loss": 0.033, "step": 26340 }, { "grad_norm": 0.41474565863609314, "learning_rate": 8.804820413445175e-05, "loss": 0.0311, "step": 26350 }, { "grad_norm": 0.2996616065502167, "learning_rate": 8.803747444109837e-05, "loss": 0.0307, "step": 26360 }, { "grad_norm": 0.25423020124435425, "learning_rate": 8.802674058802399e-05, "loss": 0.0218, "step": 26370 }, { "grad_norm": 0.4602472484111786, "learning_rate": 8.801600257640241e-05, "loss": 0.029, "step": 26380 }, { "grad_norm": 0.3905036449432373, "learning_rate": 8.800526040740795e-05, "loss": 0.0396, "step": 26390 }, { "grad_norm": 0.4800877273082733, "learning_rate": 8.799451408221535e-05, "loss": 0.0367, "step": 26400 }, { "grad_norm": 0.3068122863769531, "learning_rate": 8.798376360199982e-05, "loss": 0.0275, "step": 26410 }, { "grad_norm": 0.39914470911026, "learning_rate": 8.797300896793701e-05, "loss": 0.0252, "step": 26420 }, { "grad_norm": 0.3490867614746094, "learning_rate": 8.796225018120302e-05, "loss": 0.0388, "step": 26430 }, { "grad_norm": 0.6789354085922241, "learning_rate": 8.795148724297444e-05, "loss": 0.0253, "step": 26440 }, { "grad_norm": 0.14490565657615662, "learning_rate": 8.794072015442825e-05, "loss": 0.0291, "step": 26450 }, { "grad_norm": 0.16852982342243195, "learning_rate": 8.792994891674198e-05, "loss": 0.042, "step": 26460 }, { "grad_norm": 0.204348623752594, "learning_rate": 8.79191735310935e-05, "loss": 0.0267, "step": 26470 }, { "grad_norm": 0.15466845035552979, "learning_rate": 8.790839399866122e-05, "loss": 0.0173, "step": 26480 }, { "grad_norm": 0.32230380177497864, "learning_rate": 8.789761032062397e-05, "loss": 0.0248, "step": 26490 }, { "grad_norm": 0.4762985110282898, "learning_rate": 8.788682249816103e-05, "loss": 0.0438, "step": 26500 }, { "grad_norm": 0.5385265350341797, "learning_rate": 8.787603053245215e-05, "loss": 0.0386, "step": 26510 }, { "grad_norm": 0.3246854245662689, "learning_rate": 8.78652344246775e-05, "loss": 0.0363, "step": 26520 }, { "grad_norm": 0.3940356969833374, "learning_rate": 8.785443417601776e-05, "loss": 0.0403, "step": 26530 }, { "grad_norm": 0.32702115178108215, "learning_rate": 8.784362978765401e-05, "loss": 0.037, "step": 26540 }, { "grad_norm": 0.30908203125, "learning_rate": 8.783282126076779e-05, "loss": 0.0285, "step": 26550 }, { "grad_norm": 0.16378739476203918, "learning_rate": 8.782200859654112e-05, "loss": 0.026, "step": 26560 }, { "grad_norm": 0.32595640420913696, "learning_rate": 8.781119179615646e-05, "loss": 0.0198, "step": 26570 }, { "grad_norm": 0.2205187976360321, "learning_rate": 8.780037086079674e-05, "loss": 0.031, "step": 26580 }, { "grad_norm": 0.30099669098854065, "learning_rate": 8.778954579164527e-05, "loss": 0.0307, "step": 26590 }, { "grad_norm": 0.4905276596546173, "learning_rate": 8.777871658988588e-05, "loss": 0.0329, "step": 26600 }, { "grad_norm": 0.38909682631492615, "learning_rate": 8.776788325670285e-05, "loss": 0.0333, "step": 26610 }, { "grad_norm": 0.46768078207969666, "learning_rate": 8.775704579328089e-05, "loss": 0.0487, "step": 26620 }, { "grad_norm": 0.3067263066768646, "learning_rate": 8.774620420080517e-05, "loss": 0.0282, "step": 26630 }, { "grad_norm": 0.34632182121276855, "learning_rate": 8.773535848046131e-05, "loss": 0.0329, "step": 26640 }, { "grad_norm": 0.32265931367874146, "learning_rate": 8.772450863343538e-05, "loss": 0.0193, "step": 26650 }, { "grad_norm": 0.2702609896659851, "learning_rate": 8.77136546609139e-05, "loss": 0.0344, "step": 26660 }, { "grad_norm": 0.3177756369113922, "learning_rate": 8.770279656408385e-05, "loss": 0.0415, "step": 26670 }, { "grad_norm": 0.3953218162059784, "learning_rate": 8.769193434413265e-05, "loss": 0.0355, "step": 26680 }, { "grad_norm": 0.32837557792663574, "learning_rate": 8.76810680022482e-05, "loss": 0.0409, "step": 26690 }, { "grad_norm": 0.4538402855396271, "learning_rate": 8.767019753961878e-05, "loss": 0.0373, "step": 26700 }, { "grad_norm": 0.2580939829349518, "learning_rate": 8.765932295743321e-05, "loss": 0.0368, "step": 26710 }, { "grad_norm": 0.27784690260887146, "learning_rate": 8.764844425688068e-05, "loss": 0.0225, "step": 26720 }, { "grad_norm": 0.24636642634868622, "learning_rate": 8.763756143915092e-05, "loss": 0.0246, "step": 26730 }, { "grad_norm": 0.44253677129745483, "learning_rate": 8.7626674505434e-05, "loss": 0.0261, "step": 26740 }, { "grad_norm": 0.20920196175575256, "learning_rate": 8.761578345692053e-05, "loss": 0.0408, "step": 26750 }, { "grad_norm": 0.27867141366004944, "learning_rate": 8.760488829480156e-05, "loss": 0.0327, "step": 26760 }, { "grad_norm": 0.3426787853240967, "learning_rate": 8.759398902026854e-05, "loss": 0.0288, "step": 26770 }, { "grad_norm": 0.43775784969329834, "learning_rate": 8.758308563451339e-05, "loss": 0.0285, "step": 26780 }, { "grad_norm": 0.1951325237751007, "learning_rate": 8.75721781387285e-05, "loss": 0.0278, "step": 26790 }, { "grad_norm": 0.26791733503341675, "learning_rate": 8.75612665341067e-05, "loss": 0.0284, "step": 26800 }, { "grad_norm": 0.1474751979112625, "learning_rate": 8.755035082184126e-05, "loss": 0.0233, "step": 26810 }, { "grad_norm": 0.7000993490219116, "learning_rate": 8.753943100312592e-05, "loss": 0.0247, "step": 26820 }, { "grad_norm": 0.19804055988788605, "learning_rate": 8.752850707915484e-05, "loss": 0.0277, "step": 26830 }, { "grad_norm": 0.21444396674633026, "learning_rate": 8.751757905112264e-05, "loss": 0.035, "step": 26840 }, { "grad_norm": 0.46813470125198364, "learning_rate": 8.75066469202244e-05, "loss": 0.0377, "step": 26850 }, { "grad_norm": 0.2219519019126892, "learning_rate": 8.749571068765567e-05, "loss": 0.0357, "step": 26860 }, { "grad_norm": 0.2974408268928528, "learning_rate": 8.748477035461238e-05, "loss": 0.0331, "step": 26870 }, { "grad_norm": 0.34106388688087463, "learning_rate": 8.747382592229095e-05, "loss": 0.0349, "step": 26880 }, { "grad_norm": 0.5012449026107788, "learning_rate": 8.746287739188828e-05, "loss": 0.0316, "step": 26890 }, { "grad_norm": 0.4824337959289551, "learning_rate": 8.745192476460165e-05, "loss": 0.0368, "step": 26900 }, { "grad_norm": 0.25995391607284546, "learning_rate": 8.744096804162882e-05, "loss": 0.0247, "step": 26910 }, { "grad_norm": 0.48325613141059875, "learning_rate": 8.743000722416804e-05, "loss": 0.0339, "step": 26920 }, { "grad_norm": 0.361117959022522, "learning_rate": 8.741904231341793e-05, "loss": 0.0297, "step": 26930 }, { "grad_norm": 0.2678605914115906, "learning_rate": 8.740807331057762e-05, "loss": 0.0288, "step": 26940 }, { "grad_norm": 0.2711349427700043, "learning_rate": 8.739710021684667e-05, "loss": 0.0384, "step": 26950 }, { "grad_norm": 0.19964835047721863, "learning_rate": 8.738612303342503e-05, "loss": 0.0233, "step": 26960 }, { "grad_norm": 0.21413345634937286, "learning_rate": 8.73751417615132e-05, "loss": 0.0299, "step": 26970 }, { "grad_norm": 0.23300133645534515, "learning_rate": 8.736415640231208e-05, "loss": 0.0215, "step": 26980 }, { "grad_norm": 0.2737383246421814, "learning_rate": 8.735316695702297e-05, "loss": 0.0282, "step": 26990 }, { "grad_norm": 0.2193884402513504, "learning_rate": 8.734217342684769e-05, "loss": 0.0271, "step": 27000 }, { "grad_norm": 0.38321393728256226, "learning_rate": 8.733117581298847e-05, "loss": 0.0193, "step": 27010 }, { "grad_norm": 0.2532426714897156, "learning_rate": 8.732017411664796e-05, "loss": 0.0274, "step": 27020 }, { "grad_norm": 0.6090721487998962, "learning_rate": 8.730916833902936e-05, "loss": 0.0282, "step": 27030 }, { "grad_norm": 0.21413801610469818, "learning_rate": 8.729815848133618e-05, "loss": 0.0286, "step": 27040 }, { "grad_norm": 0.3418416976928711, "learning_rate": 8.728714454477247e-05, "loss": 0.02, "step": 27050 }, { "grad_norm": 0.4176122844219208, "learning_rate": 8.727612653054269e-05, "loss": 0.0293, "step": 27060 }, { "grad_norm": 0.33162540197372437, "learning_rate": 8.726510443985176e-05, "loss": 0.0264, "step": 27070 }, { "grad_norm": 0.34168878197669983, "learning_rate": 8.725407827390503e-05, "loss": 0.0312, "step": 27080 }, { "grad_norm": 0.31732961535453796, "learning_rate": 8.724304803390833e-05, "loss": 0.0234, "step": 27090 }, { "grad_norm": 0.20318979024887085, "learning_rate": 8.723201372106788e-05, "loss": 0.0231, "step": 27100 }, { "grad_norm": 0.33074575662612915, "learning_rate": 8.722097533659038e-05, "loss": 0.0303, "step": 27110 }, { "grad_norm": 0.2627105414867401, "learning_rate": 8.720993288168299e-05, "loss": 0.0274, "step": 27120 }, { "grad_norm": 0.5121292471885681, "learning_rate": 8.719888635755327e-05, "loss": 0.0298, "step": 27130 }, { "grad_norm": 0.1817600429058075, "learning_rate": 8.718783576540928e-05, "loss": 0.03, "step": 27140 }, { "grad_norm": 0.3126114010810852, "learning_rate": 8.717678110645948e-05, "loss": 0.0243, "step": 27150 }, { "grad_norm": 0.31274184584617615, "learning_rate": 8.716572238191279e-05, "loss": 0.0305, "step": 27160 }, { "grad_norm": 0.395373672246933, "learning_rate": 8.715465959297857e-05, "loss": 0.0312, "step": 27170 }, { "grad_norm": 0.5587466359138489, "learning_rate": 8.714359274086665e-05, "loss": 0.0347, "step": 27180 }, { "grad_norm": 0.5671409368515015, "learning_rate": 8.713252182678726e-05, "loss": 0.0473, "step": 27190 }, { "grad_norm": 0.26898062229156494, "learning_rate": 8.712144685195112e-05, "loss": 0.0331, "step": 27200 }, { "grad_norm": 0.33796024322509766, "learning_rate": 8.711036781756936e-05, "loss": 0.0301, "step": 27210 }, { "grad_norm": 0.30932191014289856, "learning_rate": 8.709928472485357e-05, "loss": 0.0229, "step": 27220 }, { "grad_norm": 0.2531389594078064, "learning_rate": 8.708819757501579e-05, "loss": 0.0288, "step": 27230 }, { "grad_norm": 0.2862240672111511, "learning_rate": 8.707710636926846e-05, "loss": 0.0314, "step": 27240 }, { "grad_norm": 0.30782878398895264, "learning_rate": 8.706601110882455e-05, "loss": 0.0352, "step": 27250 }, { "grad_norm": 0.2817276418209076, "learning_rate": 8.705491179489738e-05, "loss": 0.0272, "step": 27260 }, { "grad_norm": 0.1861751228570938, "learning_rate": 8.704380842870077e-05, "loss": 0.0245, "step": 27270 }, { "grad_norm": 0.14774762094020844, "learning_rate": 8.703270101144895e-05, "loss": 0.0292, "step": 27280 }, { "grad_norm": 0.11172333359718323, "learning_rate": 8.702158954435664e-05, "loss": 0.028, "step": 27290 }, { "grad_norm": 0.4162554144859314, "learning_rate": 8.701047402863896e-05, "loss": 0.0175, "step": 27300 }, { "grad_norm": 0.24152466654777527, "learning_rate": 8.699935446551148e-05, "loss": 0.0345, "step": 27310 }, { "grad_norm": 0.14747890830039978, "learning_rate": 8.698823085619022e-05, "loss": 0.0257, "step": 27320 }, { "grad_norm": 0.1383843868970871, "learning_rate": 8.697710320189166e-05, "loss": 0.0191, "step": 27330 }, { "grad_norm": 0.303741455078125, "learning_rate": 8.696597150383268e-05, "loss": 0.0257, "step": 27340 }, { "grad_norm": 0.26841437816619873, "learning_rate": 8.695483576323063e-05, "loss": 0.0286, "step": 27350 }, { "grad_norm": 0.6966598629951477, "learning_rate": 8.69436959813033e-05, "loss": 0.0334, "step": 27360 }, { "grad_norm": 0.2167137712240219, "learning_rate": 8.693255215926892e-05, "loss": 0.0215, "step": 27370 }, { "grad_norm": 0.28889042139053345, "learning_rate": 8.692140429834617e-05, "loss": 0.0267, "step": 27380 }, { "grad_norm": 1.0761818885803223, "learning_rate": 8.691025239975415e-05, "loss": 0.0288, "step": 27390 }, { "grad_norm": 0.21670207381248474, "learning_rate": 8.689909646471243e-05, "loss": 0.0222, "step": 27400 }, { "grad_norm": 0.2292095124721527, "learning_rate": 8.688793649444099e-05, "loss": 0.0321, "step": 27410 }, { "grad_norm": 0.2817665934562683, "learning_rate": 8.687677249016029e-05, "loss": 0.0283, "step": 27420 }, { "grad_norm": 0.4332767724990845, "learning_rate": 8.686560445309118e-05, "loss": 0.0292, "step": 27430 }, { "grad_norm": 0.2846081852912903, "learning_rate": 8.685443238445499e-05, "loss": 0.0314, "step": 27440 }, { "grad_norm": 0.34931623935699463, "learning_rate": 8.68432562854735e-05, "loss": 0.0269, "step": 27450 }, { "grad_norm": 0.20307479798793793, "learning_rate": 8.683207615736887e-05, "loss": 0.0324, "step": 27460 }, { "grad_norm": 2.555586576461792, "learning_rate": 8.682089200136379e-05, "loss": 0.0475, "step": 27470 }, { "grad_norm": 0.3483541011810303, "learning_rate": 8.680970381868132e-05, "loss": 0.0276, "step": 27480 }, { "grad_norm": 0.22977015376091003, "learning_rate": 8.679851161054498e-05, "loss": 0.0277, "step": 27490 }, { "grad_norm": 0.3226122558116913, "learning_rate": 8.678731537817873e-05, "loss": 0.0392, "step": 27500 }, { "grad_norm": 0.2714506685733795, "learning_rate": 8.677611512280697e-05, "loss": 0.0373, "step": 27510 }, { "grad_norm": 0.25107258558273315, "learning_rate": 8.676491084565457e-05, "loss": 0.0341, "step": 27520 }, { "grad_norm": 0.37484732270240784, "learning_rate": 8.675370254794678e-05, "loss": 0.0383, "step": 27530 }, { "grad_norm": 0.20935767889022827, "learning_rate": 8.674249023090935e-05, "loss": 0.0225, "step": 27540 }, { "grad_norm": 0.4142126441001892, "learning_rate": 8.673127389576843e-05, "loss": 0.0412, "step": 27550 }, { "grad_norm": 0.18614810705184937, "learning_rate": 8.67200535437506e-05, "loss": 0.0242, "step": 27560 }, { "grad_norm": 0.356145977973938, "learning_rate": 8.670882917608296e-05, "loss": 0.0315, "step": 27570 }, { "grad_norm": 0.3299182057380676, "learning_rate": 8.669760079399292e-05, "loss": 0.0269, "step": 27580 }, { "grad_norm": 0.31175968050956726, "learning_rate": 8.668636839870845e-05, "loss": 0.0251, "step": 27590 }, { "grad_norm": 0.28736814856529236, "learning_rate": 8.667513199145789e-05, "loss": 0.0445, "step": 27600 }, { "grad_norm": 0.29141801595687866, "learning_rate": 8.666389157347002e-05, "loss": 0.0317, "step": 27610 }, { "grad_norm": 0.2573643624782562, "learning_rate": 8.66526471459741e-05, "loss": 0.0269, "step": 27620 }, { "grad_norm": 0.29862964153289795, "learning_rate": 8.66413987101998e-05, "loss": 0.025, "step": 27630 }, { "grad_norm": 0.29598596692085266, "learning_rate": 8.663014626737723e-05, "loss": 0.0258, "step": 27640 }, { "grad_norm": 0.2787471413612366, "learning_rate": 8.661888981873691e-05, "loss": 0.0332, "step": 27650 }, { "grad_norm": 0.29426881670951843, "learning_rate": 8.660762936550988e-05, "loss": 0.0337, "step": 27660 }, { "grad_norm": 0.20974820852279663, "learning_rate": 8.659636490892753e-05, "loss": 0.0268, "step": 27670 }, { "grad_norm": 0.12642206251621246, "learning_rate": 8.658509645022174e-05, "loss": 0.0251, "step": 27680 }, { "grad_norm": 0.23749610781669617, "learning_rate": 8.657382399062481e-05, "loss": 0.0242, "step": 27690 }, { "grad_norm": 0.2404455989599228, "learning_rate": 8.656254753136946e-05, "loss": 0.026, "step": 27700 }, { "grad_norm": 0.3104223310947418, "learning_rate": 8.655126707368891e-05, "loss": 0.0181, "step": 27710 }, { "grad_norm": 0.24090135097503662, "learning_rate": 8.653998261881672e-05, "loss": 0.0285, "step": 27720 }, { "grad_norm": 0.442729651927948, "learning_rate": 8.652869416798699e-05, "loss": 0.0301, "step": 27730 }, { "grad_norm": 0.36752697825431824, "learning_rate": 8.651740172243417e-05, "loss": 0.0262, "step": 27740 }, { "grad_norm": 0.22854101657867432, "learning_rate": 8.65061052833932e-05, "loss": 0.0273, "step": 27750 }, { "grad_norm": 0.21215686202049255, "learning_rate": 8.649480485209945e-05, "loss": 0.0251, "step": 27760 }, { "grad_norm": 0.3136480748653412, "learning_rate": 8.64835004297887e-05, "loss": 0.0456, "step": 27770 }, { "grad_norm": 0.23003847897052765, "learning_rate": 8.64721920176972e-05, "loss": 0.0236, "step": 27780 }, { "grad_norm": 0.2826053500175476, "learning_rate": 8.646087961706164e-05, "loss": 0.0401, "step": 27790 }, { "grad_norm": 0.2000642567873001, "learning_rate": 8.644956322911908e-05, "loss": 0.0235, "step": 27800 }, { "grad_norm": 0.31869667768478394, "learning_rate": 8.643824285510709e-05, "loss": 0.0307, "step": 27810 }, { "grad_norm": 0.35388049483299255, "learning_rate": 8.642691849626364e-05, "loss": 0.0251, "step": 27820 }, { "grad_norm": 0.34603944420814514, "learning_rate": 8.641559015382717e-05, "loss": 0.0301, "step": 27830 }, { "grad_norm": 0.27686432003974915, "learning_rate": 8.640425782903649e-05, "loss": 0.0273, "step": 27840 }, { "grad_norm": 0.18198233842849731, "learning_rate": 8.639292152313091e-05, "loss": 0.0227, "step": 27850 }, { "grad_norm": 0.194660022854805, "learning_rate": 8.638158123735015e-05, "loss": 0.0273, "step": 27860 }, { "grad_norm": 0.25412318110466003, "learning_rate": 8.637023697293436e-05, "loss": 0.0238, "step": 27870 }, { "grad_norm": 0.24997222423553467, "learning_rate": 8.635888873112414e-05, "loss": 0.0468, "step": 27880 }, { "grad_norm": 0.3361288011074066, "learning_rate": 8.634753651316052e-05, "loss": 0.0277, "step": 27890 }, { "grad_norm": 0.165551096200943, "learning_rate": 8.633618032028496e-05, "loss": 0.0316, "step": 27900 }, { "grad_norm": 0.23655365407466888, "learning_rate": 8.632482015373934e-05, "loss": 0.024, "step": 27910 }, { "grad_norm": 0.295472115278244, "learning_rate": 8.6313456014766e-05, "loss": 0.0317, "step": 27920 }, { "grad_norm": 0.27873679995536804, "learning_rate": 8.630208790460771e-05, "loss": 0.0231, "step": 27930 }, { "grad_norm": 0.2746638059616089, "learning_rate": 8.629071582450768e-05, "loss": 0.0349, "step": 27940 }, { "grad_norm": 0.214517742395401, "learning_rate": 8.62793397757095e-05, "loss": 0.0284, "step": 27950 }, { "grad_norm": 0.17637485265731812, "learning_rate": 8.626795975945729e-05, "loss": 0.0386, "step": 27960 }, { "grad_norm": 0.43520262837409973, "learning_rate": 8.625657577699551e-05, "loss": 0.033, "step": 27970 }, { "grad_norm": 0.3383437991142273, "learning_rate": 8.624518782956914e-05, "loss": 0.0285, "step": 27980 }, { "grad_norm": 0.40265873074531555, "learning_rate": 8.62337959184235e-05, "loss": 0.031, "step": 27990 }, { "grad_norm": 0.250142902135849, "learning_rate": 8.622240004480441e-05, "loss": 0.0349, "step": 28000 }, { "grad_norm": 0.25588369369506836, "learning_rate": 8.621100020995814e-05, "loss": 0.0309, "step": 28010 }, { "grad_norm": 0.42695489525794983, "learning_rate": 8.619959641513132e-05, "loss": 0.0526, "step": 28020 }, { "grad_norm": 0.13674531877040863, "learning_rate": 8.618818866157105e-05, "loss": 0.0274, "step": 28030 }, { "grad_norm": 0.25818273425102234, "learning_rate": 8.617677695052487e-05, "loss": 0.0314, "step": 28040 }, { "grad_norm": 0.25272807478904724, "learning_rate": 8.616536128324078e-05, "loss": 0.0326, "step": 28050 }, { "grad_norm": 0.3422456383705139, "learning_rate": 8.615394166096712e-05, "loss": 0.0312, "step": 28060 }, { "grad_norm": 0.329073965549469, "learning_rate": 8.614251808495279e-05, "loss": 0.0306, "step": 28070 }, { "grad_norm": 0.2304425984621048, "learning_rate": 8.6131090556447e-05, "loss": 0.0253, "step": 28080 }, { "grad_norm": 0.3950672745704651, "learning_rate": 8.611965907669947e-05, "loss": 0.0382, "step": 28090 }, { "grad_norm": 0.28420495986938477, "learning_rate": 8.610822364696034e-05, "loss": 0.0315, "step": 28100 }, { "grad_norm": 0.20613670349121094, "learning_rate": 8.609678426848015e-05, "loss": 0.0255, "step": 28110 }, { "grad_norm": 0.38340896368026733, "learning_rate": 8.60853409425099e-05, "loss": 0.0359, "step": 28120 }, { "grad_norm": 0.4590196907520294, "learning_rate": 8.607389367030104e-05, "loss": 0.0337, "step": 28130 }, { "grad_norm": 0.3548862338066101, "learning_rate": 8.606244245310538e-05, "loss": 0.0225, "step": 28140 }, { "grad_norm": 0.33339977264404297, "learning_rate": 8.605098729217525e-05, "loss": 0.0422, "step": 28150 }, { "grad_norm": 0.17269468307495117, "learning_rate": 8.603952818876335e-05, "loss": 0.0246, "step": 28160 }, { "grad_norm": 0.4885573089122772, "learning_rate": 8.602806514412281e-05, "loss": 0.0226, "step": 28170 }, { "grad_norm": 0.6082981824874878, "learning_rate": 8.601659815950726e-05, "loss": 0.0472, "step": 28180 }, { "grad_norm": 0.22498148679733276, "learning_rate": 8.600512723617067e-05, "loss": 0.031, "step": 28190 }, { "grad_norm": 0.5322617888450623, "learning_rate": 8.59936523753675e-05, "loss": 0.0316, "step": 28200 }, { "grad_norm": 0.1449000984430313, "learning_rate": 8.598217357835264e-05, "loss": 0.0383, "step": 28210 }, { "grad_norm": 0.21058280766010284, "learning_rate": 8.597069084638135e-05, "loss": 0.0273, "step": 28220 }, { "grad_norm": 0.22774949669837952, "learning_rate": 8.595920418070939e-05, "loss": 0.0294, "step": 28230 }, { "grad_norm": 0.4052179753780365, "learning_rate": 8.594771358259295e-05, "loss": 0.0349, "step": 28240 }, { "grad_norm": 0.23453666269779205, "learning_rate": 8.593621905328858e-05, "loss": 0.0301, "step": 28250 }, { "grad_norm": 0.3229656517505646, "learning_rate": 8.592472059405333e-05, "loss": 0.0332, "step": 28260 }, { "grad_norm": 0.11809705942869186, "learning_rate": 8.591321820614464e-05, "loss": 0.0231, "step": 28270 }, { "grad_norm": 0.2427421659231186, "learning_rate": 8.590171189082041e-05, "loss": 0.0273, "step": 28280 }, { "grad_norm": 0.30344969034194946, "learning_rate": 8.589020164933894e-05, "loss": 0.0318, "step": 28290 }, { "grad_norm": 0.18822821974754333, "learning_rate": 8.587868748295898e-05, "loss": 0.0275, "step": 28300 }, { "grad_norm": 0.30935031175613403, "learning_rate": 8.586716939293971e-05, "loss": 0.0327, "step": 28310 }, { "grad_norm": 0.25766557455062866, "learning_rate": 8.58556473805407e-05, "loss": 0.0268, "step": 28320 }, { "grad_norm": 0.22990518808364868, "learning_rate": 8.584412144702202e-05, "loss": 0.0297, "step": 28330 }, { "grad_norm": 0.42911389470100403, "learning_rate": 8.58325915936441e-05, "loss": 0.0323, "step": 28340 }, { "grad_norm": 0.2461366355419159, "learning_rate": 8.582105782166783e-05, "loss": 0.026, "step": 28350 }, { "grad_norm": 0.2643886208534241, "learning_rate": 8.580952013235455e-05, "loss": 0.0246, "step": 28360 }, { "grad_norm": 0.11589518189430237, "learning_rate": 8.579797852696596e-05, "loss": 0.0177, "step": 28370 }, { "grad_norm": 0.20199209451675415, "learning_rate": 8.578643300676428e-05, "loss": 0.0193, "step": 28380 }, { "grad_norm": 0.13004061579704285, "learning_rate": 8.577488357301209e-05, "loss": 0.0244, "step": 28390 }, { "grad_norm": 0.22085990011692047, "learning_rate": 8.576333022697242e-05, "loss": 0.0228, "step": 28400 }, { "grad_norm": 0.25413069128990173, "learning_rate": 8.575177296990873e-05, "loss": 0.0243, "step": 28410 }, { "grad_norm": 0.1599123477935791, "learning_rate": 8.574021180308489e-05, "loss": 0.0303, "step": 28420 }, { "grad_norm": 0.35146427154541016, "learning_rate": 8.572864672776523e-05, "loss": 0.0251, "step": 28430 }, { "grad_norm": 0.1942528486251831, "learning_rate": 8.571707774521447e-05, "loss": 0.0241, "step": 28440 }, { "grad_norm": 0.2235756814479828, "learning_rate": 8.57055048566978e-05, "loss": 0.0237, "step": 28450 }, { "grad_norm": 0.2551368474960327, "learning_rate": 8.569392806348078e-05, "loss": 0.0281, "step": 28460 }, { "grad_norm": 0.3149469792842865, "learning_rate": 8.568234736682947e-05, "loss": 0.0364, "step": 28470 }, { "grad_norm": 0.3188186585903168, "learning_rate": 8.567076276801029e-05, "loss": 0.034, "step": 28480 }, { "grad_norm": 0.302453875541687, "learning_rate": 8.565917426829013e-05, "loss": 0.0291, "step": 28490 }, { "grad_norm": 0.23138073086738586, "learning_rate": 8.564758186893628e-05, "loss": 0.022, "step": 28500 }, { "grad_norm": 0.4567602574825287, "learning_rate": 8.563598557121649e-05, "loss": 0.0351, "step": 28510 }, { "grad_norm": 0.4921127259731293, "learning_rate": 8.562438537639888e-05, "loss": 0.0311, "step": 28520 }, { "grad_norm": 0.24271757900714874, "learning_rate": 8.561278128575206e-05, "loss": 0.0227, "step": 28530 }, { "grad_norm": 0.30112424492836, "learning_rate": 8.5601173300545e-05, "loss": 0.0235, "step": 28540 }, { "grad_norm": 0.28790053725242615, "learning_rate": 8.558956142204717e-05, "loss": 0.0241, "step": 28550 }, { "grad_norm": 0.27876925468444824, "learning_rate": 8.55779456515284e-05, "loss": 0.026, "step": 28560 }, { "grad_norm": 0.18327803909778595, "learning_rate": 8.556632599025898e-05, "loss": 0.0301, "step": 28570 }, { "grad_norm": 0.24099408090114594, "learning_rate": 8.555470243950964e-05, "loss": 0.0337, "step": 28580 }, { "grad_norm": 0.1852194219827652, "learning_rate": 8.554307500055148e-05, "loss": 0.0256, "step": 28590 }, { "grad_norm": 0.4239217936992645, "learning_rate": 8.553144367465609e-05, "loss": 0.0301, "step": 28600 }, { "grad_norm": 0.35939452052116394, "learning_rate": 8.551980846309544e-05, "loss": 0.0329, "step": 28610 }, { "grad_norm": 0.1851351112127304, "learning_rate": 8.550816936714193e-05, "loss": 0.0283, "step": 28620 }, { "grad_norm": 0.34300220012664795, "learning_rate": 8.549652638806841e-05, "loss": 0.0275, "step": 28630 }, { "grad_norm": 0.2075709104537964, "learning_rate": 8.548487952714812e-05, "loss": 0.0191, "step": 28640 }, { "grad_norm": 0.199689120054245, "learning_rate": 8.547322878565478e-05, "loss": 0.0285, "step": 28650 }, { "grad_norm": 0.4455684423446655, "learning_rate": 8.546157416486245e-05, "loss": 0.0346, "step": 28660 }, { "grad_norm": 0.1568162739276886, "learning_rate": 8.54499156660457e-05, "loss": 0.0187, "step": 28670 }, { "grad_norm": 0.21330544352531433, "learning_rate": 8.543825329047947e-05, "loss": 0.0217, "step": 28680 }, { "grad_norm": 0.1640273481607437, "learning_rate": 8.542658703943913e-05, "loss": 0.0288, "step": 28690 }, { "grad_norm": 0.22903697192668915, "learning_rate": 8.541491691420051e-05, "loss": 0.0286, "step": 28700 }, { "grad_norm": 0.19476677477359772, "learning_rate": 8.54032429160398e-05, "loss": 0.019, "step": 28710 }, { "grad_norm": 0.2541802227497101, "learning_rate": 8.539156504623369e-05, "loss": 0.0404, "step": 28720 }, { "grad_norm": 0.17906583845615387, "learning_rate": 8.537988330605923e-05, "loss": 0.0343, "step": 28730 }, { "grad_norm": 0.27341949939727783, "learning_rate": 8.536819769679393e-05, "loss": 0.0193, "step": 28740 }, { "grad_norm": 0.3738705515861511, "learning_rate": 8.53565082197157e-05, "loss": 0.0249, "step": 28750 }, { "grad_norm": 0.23099029064178467, "learning_rate": 8.534481487610289e-05, "loss": 0.0311, "step": 28760 }, { "grad_norm": 0.32009968161582947, "learning_rate": 8.533311766723428e-05, "loss": 0.0189, "step": 28770 }, { "grad_norm": 0.48172253370285034, "learning_rate": 8.532141659438901e-05, "loss": 0.0301, "step": 28780 }, { "grad_norm": 0.24980191886425018, "learning_rate": 8.530971165884675e-05, "loss": 0.0293, "step": 28790 }, { "grad_norm": 0.27344435453414917, "learning_rate": 8.529800286188752e-05, "loss": 0.0206, "step": 28800 }, { "grad_norm": 0.15418359637260437, "learning_rate": 8.528629020479175e-05, "loss": 0.0273, "step": 28810 }, { "grad_norm": 0.19169649481773376, "learning_rate": 8.527457368884033e-05, "loss": 0.0308, "step": 28820 }, { "grad_norm": 0.2771220803260803, "learning_rate": 8.526285331531458e-05, "loss": 0.0211, "step": 28830 }, { "grad_norm": 0.3104150593280792, "learning_rate": 8.525112908549621e-05, "loss": 0.0271, "step": 28840 }, { "grad_norm": 0.2326136976480484, "learning_rate": 8.523940100066735e-05, "loss": 0.0295, "step": 28850 }, { "grad_norm": 0.24771419167518616, "learning_rate": 8.52276690621106e-05, "loss": 0.0221, "step": 28860 }, { "grad_norm": 0.4207969903945923, "learning_rate": 8.521593327110889e-05, "loss": 0.0281, "step": 28870 }, { "grad_norm": 0.5788025856018066, "learning_rate": 8.520419362894569e-05, "loss": 0.0432, "step": 28880 }, { "grad_norm": 0.3031924366950989, "learning_rate": 8.51924501369048e-05, "loss": 0.025, "step": 28890 }, { "grad_norm": 0.4422287344932556, "learning_rate": 8.518070279627047e-05, "loss": 0.0328, "step": 28900 }, { "grad_norm": 0.16560368239879608, "learning_rate": 8.516895160832737e-05, "loss": 0.0316, "step": 28910 }, { "grad_norm": 0.26467272639274597, "learning_rate": 8.515719657436061e-05, "loss": 0.034, "step": 28920 }, { "grad_norm": 0.38142186403274536, "learning_rate": 8.514543769565568e-05, "loss": 0.0282, "step": 28930 }, { "grad_norm": 0.4181654453277588, "learning_rate": 8.513367497349853e-05, "loss": 0.0288, "step": 28940 }, { "grad_norm": 0.3864595592021942, "learning_rate": 8.51219084091755e-05, "loss": 0.0235, "step": 28950 }, { "grad_norm": 0.26243820786476135, "learning_rate": 8.511013800397338e-05, "loss": 0.0157, "step": 28960 }, { "grad_norm": 0.16402176022529602, "learning_rate": 8.509836375917937e-05, "loss": 0.0246, "step": 28970 }, { "grad_norm": 0.1587701439857483, "learning_rate": 8.508658567608104e-05, "loss": 0.0205, "step": 28980 }, { "grad_norm": 0.32119786739349365, "learning_rate": 8.507480375596647e-05, "loss": 0.0292, "step": 28990 }, { "grad_norm": 0.26277390122413635, "learning_rate": 8.506301800012408e-05, "loss": 0.0261, "step": 29000 }, { "grad_norm": 0.38224583864212036, "learning_rate": 8.505122840984278e-05, "loss": 0.0343, "step": 29010 }, { "grad_norm": 0.17019033432006836, "learning_rate": 8.503943498641182e-05, "loss": 0.0297, "step": 29020 }, { "grad_norm": 0.16299954056739807, "learning_rate": 8.502763773112095e-05, "loss": 0.027, "step": 29030 }, { "grad_norm": 0.24261538684368134, "learning_rate": 8.501583664526026e-05, "loss": 0.0359, "step": 29040 }, { "grad_norm": 0.2889825403690338, "learning_rate": 8.500403173012032e-05, "loss": 0.0333, "step": 29050 }, { "grad_norm": 0.48432186245918274, "learning_rate": 8.499222298699211e-05, "loss": 0.0302, "step": 29060 }, { "grad_norm": 0.18272387981414795, "learning_rate": 8.498041041716701e-05, "loss": 0.031, "step": 29070 }, { "grad_norm": 0.21457041800022125, "learning_rate": 8.496859402193681e-05, "loss": 0.0261, "step": 29080 }, { "grad_norm": 0.3235245645046234, "learning_rate": 8.495677380259374e-05, "loss": 0.0285, "step": 29090 }, { "grad_norm": 0.3865925967693329, "learning_rate": 8.494494976043045e-05, "loss": 0.0344, "step": 29100 }, { "grad_norm": 0.24812181293964386, "learning_rate": 8.493312189673998e-05, "loss": 0.0297, "step": 29110 }, { "grad_norm": 0.24297723174095154, "learning_rate": 8.492129021281584e-05, "loss": 0.0327, "step": 29120 }, { "grad_norm": 0.17665618658065796, "learning_rate": 8.490945470995188e-05, "loss": 0.023, "step": 29130 }, { "grad_norm": 0.3314448893070221, "learning_rate": 8.489761538944247e-05, "loss": 0.0318, "step": 29140 }, { "grad_norm": 0.3443554639816284, "learning_rate": 8.48857722525823e-05, "loss": 0.028, "step": 29150 }, { "grad_norm": 0.46798574924468994, "learning_rate": 8.487392530066652e-05, "loss": 0.0346, "step": 29160 }, { "grad_norm": 0.18732979893684387, "learning_rate": 8.486207453499069e-05, "loss": 0.0279, "step": 29170 }, { "grad_norm": 0.45778489112854004, "learning_rate": 8.485021995685082e-05, "loss": 0.032, "step": 29180 }, { "grad_norm": 0.29332610964775085, "learning_rate": 8.483836156754328e-05, "loss": 0.0229, "step": 29190 }, { "grad_norm": 0.19546131789684296, "learning_rate": 8.482649936836491e-05, "loss": 0.024, "step": 29200 }, { "grad_norm": 0.3803628385066986, "learning_rate": 8.481463336061293e-05, "loss": 0.0232, "step": 29210 }, { "grad_norm": 0.18092040717601776, "learning_rate": 8.480276354558496e-05, "loss": 0.0289, "step": 29220 }, { "grad_norm": 0.245576873421669, "learning_rate": 8.479088992457913e-05, "loss": 0.0271, "step": 29230 }, { "grad_norm": 0.2740892171859741, "learning_rate": 8.477901249889387e-05, "loss": 0.0398, "step": 29240 }, { "grad_norm": 0.37657439708709717, "learning_rate": 8.47671312698281e-05, "loss": 0.0264, "step": 29250 }, { "grad_norm": 0.1370379477739334, "learning_rate": 8.475524623868112e-05, "loss": 0.0185, "step": 29260 }, { "grad_norm": 0.229952871799469, "learning_rate": 8.474335740675266e-05, "loss": 0.0263, "step": 29270 }, { "grad_norm": 0.30121564865112305, "learning_rate": 8.473146477534289e-05, "loss": 0.0266, "step": 29280 }, { "grad_norm": 0.4327038526535034, "learning_rate": 8.471956834575232e-05, "loss": 0.0217, "step": 29290 }, { "grad_norm": 0.41607746481895447, "learning_rate": 8.470766811928197e-05, "loss": 0.0276, "step": 29300 }, { "grad_norm": 0.26011478900909424, "learning_rate": 8.469576409723323e-05, "loss": 0.027, "step": 29310 }, { "grad_norm": 0.47537264227867126, "learning_rate": 8.468385628090788e-05, "loss": 0.0236, "step": 29320 }, { "grad_norm": 0.2124360352754593, "learning_rate": 8.467194467160815e-05, "loss": 0.0308, "step": 29330 }, { "grad_norm": 0.20216453075408936, "learning_rate": 8.466002927063667e-05, "loss": 0.021, "step": 29340 }, { "grad_norm": 0.3015938103199005, "learning_rate": 8.464811007929651e-05, "loss": 0.0324, "step": 29350 }, { "grad_norm": 0.35551831126213074, "learning_rate": 8.463618709889114e-05, "loss": 0.0265, "step": 29360 }, { "grad_norm": 0.26669856905937195, "learning_rate": 8.462426033072442e-05, "loss": 0.0281, "step": 29370 }, { "grad_norm": 0.1969028264284134, "learning_rate": 8.461232977610061e-05, "loss": 0.0253, "step": 29380 }, { "grad_norm": 0.3676728904247284, "learning_rate": 8.46003954363245e-05, "loss": 0.0177, "step": 29390 }, { "grad_norm": 0.1642036885023117, "learning_rate": 8.458845731270115e-05, "loss": 0.0279, "step": 29400 }, { "grad_norm": 0.282147079706192, "learning_rate": 8.45765154065361e-05, "loss": 0.024, "step": 29410 }, { "grad_norm": 0.14233481884002686, "learning_rate": 8.456456971913532e-05, "loss": 0.0166, "step": 29420 }, { "grad_norm": 0.3083580732345581, "learning_rate": 8.455262025180517e-05, "loss": 0.0278, "step": 29430 }, { "grad_norm": 0.18425729870796204, "learning_rate": 8.454066700585242e-05, "loss": 0.0226, "step": 29440 }, { "grad_norm": 0.15084166824817657, "learning_rate": 8.452870998258423e-05, "loss": 0.0281, "step": 29450 }, { "grad_norm": 0.21497958898544312, "learning_rate": 8.451674918330825e-05, "loss": 0.0167, "step": 29460 }, { "grad_norm": 0.2060830295085907, "learning_rate": 8.450478460933246e-05, "loss": 0.0281, "step": 29470 }, { "grad_norm": 0.32998037338256836, "learning_rate": 8.449281626196532e-05, "loss": 0.0343, "step": 29480 }, { "grad_norm": 0.23966878652572632, "learning_rate": 8.448084414251564e-05, "loss": 0.0228, "step": 29490 }, { "grad_norm": 0.4503173232078552, "learning_rate": 8.446886825229271e-05, "loss": 0.0275, "step": 29500 }, { "grad_norm": 0.27049610018730164, "learning_rate": 8.445688859260615e-05, "loss": 0.0234, "step": 29510 }, { "grad_norm": 0.3187892436981201, "learning_rate": 8.444490516476606e-05, "loss": 0.0232, "step": 29520 }, { "grad_norm": 0.22903192043304443, "learning_rate": 8.443291797008293e-05, "loss": 0.0299, "step": 29530 }, { "grad_norm": 0.34693294763565063, "learning_rate": 8.442092700986765e-05, "loss": 0.0385, "step": 29540 }, { "grad_norm": 0.3665611445903778, "learning_rate": 8.440893228543156e-05, "loss": 0.0309, "step": 29550 }, { "grad_norm": 0.23078188300132751, "learning_rate": 8.439693379808638e-05, "loss": 0.0303, "step": 29560 }, { "grad_norm": 0.15117749571800232, "learning_rate": 8.43849315491442e-05, "loss": 0.025, "step": 29570 }, { "grad_norm": 0.182996928691864, "learning_rate": 8.437292553991763e-05, "loss": 0.0268, "step": 29580 }, { "grad_norm": 0.31594306230545044, "learning_rate": 8.436091577171959e-05, "loss": 0.0267, "step": 29590 }, { "grad_norm": 0.27767622470855713, "learning_rate": 8.434890224586347e-05, "loss": 0.0198, "step": 29600 }, { "grad_norm": 0.27464866638183594, "learning_rate": 8.433688496366303e-05, "loss": 0.0214, "step": 29610 }, { "grad_norm": 0.378031849861145, "learning_rate": 8.432486392643248e-05, "loss": 0.0297, "step": 29620 }, { "grad_norm": 0.21039526164531708, "learning_rate": 8.431283913548643e-05, "loss": 0.0312, "step": 29630 }, { "grad_norm": 0.38016095757484436, "learning_rate": 8.430081059213985e-05, "loss": 0.023, "step": 29640 }, { "grad_norm": 0.3067311644554138, "learning_rate": 8.428877829770823e-05, "loss": 0.0306, "step": 29650 }, { "grad_norm": 0.296281635761261, "learning_rate": 8.427674225350735e-05, "loss": 0.0297, "step": 29660 }, { "grad_norm": 0.33605867624282837, "learning_rate": 8.426470246085347e-05, "loss": 0.0315, "step": 29670 }, { "grad_norm": 0.18399971723556519, "learning_rate": 8.425265892106324e-05, "loss": 0.0281, "step": 29680 }, { "grad_norm": 0.2850324809551239, "learning_rate": 8.424061163545374e-05, "loss": 0.0313, "step": 29690 }, { "grad_norm": 0.2300293743610382, "learning_rate": 8.422856060534243e-05, "loss": 0.0241, "step": 29700 }, { "grad_norm": 0.35172927379608154, "learning_rate": 8.421650583204718e-05, "loss": 0.0279, "step": 29710 }, { "grad_norm": 0.33782583475112915, "learning_rate": 8.420444731688633e-05, "loss": 0.0303, "step": 29720 }, { "grad_norm": 0.20597194135189056, "learning_rate": 8.419238506117852e-05, "loss": 0.0346, "step": 29730 }, { "grad_norm": 0.24513448774814606, "learning_rate": 8.418031906624289e-05, "loss": 0.0264, "step": 29740 }, { "grad_norm": 0.4426068961620331, "learning_rate": 8.416824933339898e-05, "loss": 0.0306, "step": 29750 }, { "grad_norm": 0.2725566625595093, "learning_rate": 8.415617586396667e-05, "loss": 0.0225, "step": 29760 }, { "grad_norm": 0.23473554849624634, "learning_rate": 8.414409865926632e-05, "loss": 0.0245, "step": 29770 }, { "grad_norm": 0.33862972259521484, "learning_rate": 8.413201772061867e-05, "loss": 0.0287, "step": 29780 }, { "grad_norm": 0.2760235667228699, "learning_rate": 8.411993304934488e-05, "loss": 0.0182, "step": 29790 }, { "grad_norm": 0.09776981174945831, "learning_rate": 8.410784464676654e-05, "loss": 0.021, "step": 29800 }, { "grad_norm": 0.24429744482040405, "learning_rate": 8.409575251420556e-05, "loss": 0.022, "step": 29810 }, { "grad_norm": 0.20766308903694153, "learning_rate": 8.408365665298435e-05, "loss": 0.0305, "step": 29820 }, { "grad_norm": 0.28591036796569824, "learning_rate": 8.40715570644257e-05, "loss": 0.0393, "step": 29830 }, { "grad_norm": 0.27942904829978943, "learning_rate": 8.40594537498528e-05, "loss": 0.0428, "step": 29840 }, { "grad_norm": 0.42076370120048523, "learning_rate": 8.404734671058924e-05, "loss": 0.0324, "step": 29850 }, { "grad_norm": 0.2983206510543823, "learning_rate": 8.403523594795902e-05, "loss": 0.0485, "step": 29860 }, { "grad_norm": 0.4408041536808014, "learning_rate": 8.402312146328659e-05, "loss": 0.0291, "step": 29870 }, { "grad_norm": 0.282182902097702, "learning_rate": 8.401100325789675e-05, "loss": 0.0309, "step": 29880 }, { "grad_norm": 0.3747583329677582, "learning_rate": 8.399888133311472e-05, "loss": 0.0315, "step": 29890 }, { "grad_norm": 0.38358500599861145, "learning_rate": 8.398675569026613e-05, "loss": 0.0351, "step": 29900 }, { "grad_norm": 0.22818273305892944, "learning_rate": 8.397462633067705e-05, "loss": 0.0255, "step": 29910 }, { "grad_norm": 0.17104443907737732, "learning_rate": 8.396249325567392e-05, "loss": 0.0382, "step": 29920 }, { "grad_norm": 0.2782660722732544, "learning_rate": 8.395035646658357e-05, "loss": 0.0351, "step": 29930 }, { "grad_norm": 0.20937447249889374, "learning_rate": 8.39382159647333e-05, "loss": 0.0284, "step": 29940 }, { "grad_norm": 0.3347770571708679, "learning_rate": 8.392607175145075e-05, "loss": 0.0254, "step": 29950 }, { "grad_norm": 0.34349754452705383, "learning_rate": 8.3913923828064e-05, "loss": 0.0317, "step": 29960 }, { "grad_norm": 0.16904206573963165, "learning_rate": 8.390177219590152e-05, "loss": 0.0224, "step": 29970 }, { "grad_norm": 0.46510621905326843, "learning_rate": 8.388961685629222e-05, "loss": 0.0343, "step": 29980 }, { "grad_norm": 0.26827743649482727, "learning_rate": 8.387745781056536e-05, "loss": 0.0283, "step": 29990 }, { "grad_norm": 0.3541405498981476, "learning_rate": 8.386529506005065e-05, "loss": 0.0258, "step": 30000 }, { "grad_norm": 0.29316145181655884, "learning_rate": 8.38531286060782e-05, "loss": 0.0288, "step": 30010 }, { "grad_norm": 0.3084030747413635, "learning_rate": 8.384095844997849e-05, "loss": 0.0248, "step": 30020 }, { "grad_norm": 0.3178403675556183, "learning_rate": 8.382878459308245e-05, "loss": 0.0184, "step": 30030 }, { "grad_norm": 0.4603712856769562, "learning_rate": 8.381660703672138e-05, "loss": 0.0249, "step": 30040 }, { "grad_norm": 0.37217360734939575, "learning_rate": 8.380442578222702e-05, "loss": 0.039, "step": 30050 }, { "grad_norm": 0.11711666733026505, "learning_rate": 8.379224083093146e-05, "loss": 0.02, "step": 30060 }, { "grad_norm": 0.6400112509727478, "learning_rate": 8.378005218416727e-05, "loss": 0.0343, "step": 30070 }, { "grad_norm": 0.19751693308353424, "learning_rate": 8.376785984326735e-05, "loss": 0.0317, "step": 30080 }, { "grad_norm": 0.23742252588272095, "learning_rate": 8.375566380956506e-05, "loss": 0.0238, "step": 30090 }, { "grad_norm": 0.39948761463165283, "learning_rate": 8.374346408439411e-05, "loss": 0.042, "step": 30100 }, { "grad_norm": 0.7722305655479431, "learning_rate": 8.373126066908868e-05, "loss": 0.0472, "step": 30110 }, { "grad_norm": 0.2901381552219391, "learning_rate": 8.371905356498326e-05, "loss": 0.0282, "step": 30120 }, { "grad_norm": 0.21700528264045715, "learning_rate": 8.370684277341288e-05, "loss": 0.0292, "step": 30130 }, { "grad_norm": 0.4816230833530426, "learning_rate": 8.369462829571282e-05, "loss": 0.0322, "step": 30140 }, { "grad_norm": 0.1890438050031662, "learning_rate": 8.36824101332189e-05, "loss": 0.0297, "step": 30150 }, { "grad_norm": 0.23941396176815033, "learning_rate": 8.367018828726721e-05, "loss": 0.026, "step": 30160 }, { "grad_norm": 0.24356023967266083, "learning_rate": 8.365796275919438e-05, "loss": 0.0212, "step": 30170 }, { "grad_norm": 0.17010968923568726, "learning_rate": 8.364573355033734e-05, "loss": 0.0293, "step": 30180 }, { "grad_norm": 0.2005007416009903, "learning_rate": 8.363350066203346e-05, "loss": 0.026, "step": 30190 }, { "grad_norm": 0.2551596164703369, "learning_rate": 8.362126409562053e-05, "loss": 0.033, "step": 30200 }, { "grad_norm": 0.3816775679588318, "learning_rate": 8.360902385243667e-05, "loss": 0.0283, "step": 30210 }, { "grad_norm": 0.2931765615940094, "learning_rate": 8.359677993382052e-05, "loss": 0.0236, "step": 30220 }, { "grad_norm": 0.293112188577652, "learning_rate": 8.358453234111103e-05, "loss": 0.03, "step": 30230 }, { "grad_norm": 0.3664294183254242, "learning_rate": 8.357228107564756e-05, "loss": 0.0422, "step": 30240 }, { "grad_norm": 0.2761879861354828, "learning_rate": 8.356002613876993e-05, "loss": 0.0231, "step": 30250 }, { "grad_norm": 0.2635275721549988, "learning_rate": 8.35477675318183e-05, "loss": 0.0151, "step": 30260 }, { "grad_norm": 0.30471107363700867, "learning_rate": 8.353550525613323e-05, "loss": 0.0358, "step": 30270 }, { "grad_norm": 0.14425022900104523, "learning_rate": 8.352323931305572e-05, "loss": 0.0382, "step": 30280 }, { "grad_norm": 0.31385594606399536, "learning_rate": 8.351096970392717e-05, "loss": 0.0299, "step": 30290 }, { "grad_norm": 0.3896253705024719, "learning_rate": 8.349869643008937e-05, "loss": 0.0273, "step": 30300 }, { "grad_norm": 0.3910677433013916, "learning_rate": 8.348641949288449e-05, "loss": 0.0343, "step": 30310 }, { "grad_norm": 0.24103249609470367, "learning_rate": 8.34741388936551e-05, "loss": 0.0345, "step": 30320 }, { "grad_norm": 0.21227048337459564, "learning_rate": 8.346185463374423e-05, "loss": 0.0323, "step": 30330 }, { "grad_norm": 0.18280020356178284, "learning_rate": 8.344956671449524e-05, "loss": 0.0326, "step": 30340 }, { "grad_norm": 0.2868838310241699, "learning_rate": 8.343727513725192e-05, "loss": 0.0247, "step": 30350 }, { "grad_norm": 0.2267223447561264, "learning_rate": 8.342497990335847e-05, "loss": 0.0362, "step": 30360 }, { "grad_norm": 0.2878912687301636, "learning_rate": 8.341268101415946e-05, "loss": 0.0332, "step": 30370 }, { "grad_norm": 0.33492064476013184, "learning_rate": 8.34003784709999e-05, "loss": 0.0302, "step": 30380 }, { "grad_norm": 0.2918737530708313, "learning_rate": 8.338807227522516e-05, "loss": 0.0262, "step": 30390 }, { "grad_norm": 0.27270448207855225, "learning_rate": 8.337576242818103e-05, "loss": 0.0263, "step": 30400 }, { "grad_norm": 0.19180159270763397, "learning_rate": 8.336344893121372e-05, "loss": 0.015, "step": 30410 }, { "grad_norm": 0.18268676102161407, "learning_rate": 8.335113178566977e-05, "loss": 0.0244, "step": 30420 }, { "grad_norm": 0.2578413486480713, "learning_rate": 8.33388109928962e-05, "loss": 0.0364, "step": 30430 }, { "grad_norm": 0.22047138214111328, "learning_rate": 8.33264865542404e-05, "loss": 0.022, "step": 30440 }, { "grad_norm": 0.22214944660663605, "learning_rate": 8.331415847105013e-05, "loss": 0.0302, "step": 30450 }, { "grad_norm": 0.25546732544898987, "learning_rate": 8.330182674467357e-05, "loss": 0.027, "step": 30460 }, { "grad_norm": 0.2162550538778305, "learning_rate": 8.32894913764593e-05, "loss": 0.0273, "step": 30470 }, { "grad_norm": 0.1625831127166748, "learning_rate": 8.327715236775633e-05, "loss": 0.0189, "step": 30480 }, { "grad_norm": 0.39753445982933044, "learning_rate": 8.326480971991398e-05, "loss": 0.027, "step": 30490 }, { "grad_norm": 0.19810770452022552, "learning_rate": 8.325246343428206e-05, "loss": 0.0268, "step": 30500 }, { "grad_norm": 0.22552894055843353, "learning_rate": 8.324011351221072e-05, "loss": 0.0231, "step": 30510 }, { "grad_norm": 0.3275967538356781, "learning_rate": 8.322775995505057e-05, "loss": 0.0269, "step": 30520 }, { "grad_norm": 0.32440653443336487, "learning_rate": 8.321540276415254e-05, "loss": 0.0214, "step": 30530 }, { "grad_norm": 0.3749570846557617, "learning_rate": 8.320304194086798e-05, "loss": 0.0315, "step": 30540 }, { "grad_norm": 0.17098455131053925, "learning_rate": 8.31906774865487e-05, "loss": 0.0233, "step": 30550 }, { "grad_norm": 0.3052978217601776, "learning_rate": 8.317830940254682e-05, "loss": 0.0254, "step": 30560 }, { "grad_norm": 0.30845651030540466, "learning_rate": 8.316593769021491e-05, "loss": 0.0359, "step": 30570 }, { "grad_norm": 0.3524434566497803, "learning_rate": 8.315356235090592e-05, "loss": 0.0261, "step": 30580 }, { "grad_norm": 0.215637668967247, "learning_rate": 8.314118338597319e-05, "loss": 0.0317, "step": 30590 }, { "grad_norm": 0.25448545813560486, "learning_rate": 8.312880079677048e-05, "loss": 0.0214, "step": 30600 }, { "grad_norm": 0.6152940392494202, "learning_rate": 8.311641458465191e-05, "loss": 0.0262, "step": 30610 }, { "grad_norm": 0.28300756216049194, "learning_rate": 8.310402475097205e-05, "loss": 0.023, "step": 30620 }, { "grad_norm": 0.29742157459259033, "learning_rate": 8.309163129708581e-05, "loss": 0.0253, "step": 30630 }, { "grad_norm": 0.1917058378458023, "learning_rate": 8.307923422434852e-05, "loss": 0.0247, "step": 30640 }, { "grad_norm": 0.28412124514579773, "learning_rate": 8.30668335341159e-05, "loss": 0.0297, "step": 30650 }, { "grad_norm": 0.22445006668567657, "learning_rate": 8.305442922774408e-05, "loss": 0.023, "step": 30660 }, { "grad_norm": 0.2934129536151886, "learning_rate": 8.304202130658959e-05, "loss": 0.0284, "step": 30670 }, { "grad_norm": 0.22666263580322266, "learning_rate": 8.302960977200931e-05, "loss": 0.0195, "step": 30680 }, { "grad_norm": 0.472843736410141, "learning_rate": 8.301719462536058e-05, "loss": 0.0232, "step": 30690 }, { "grad_norm": 0.3705274164676666, "learning_rate": 8.300477586800108e-05, "loss": 0.0272, "step": 30700 }, { "grad_norm": 0.568638265132904, "learning_rate": 8.299235350128892e-05, "loss": 0.0268, "step": 30710 }, { "grad_norm": 0.5883612632751465, "learning_rate": 8.297992752658258e-05, "loss": 0.0233, "step": 30720 }, { "grad_norm": 0.21589136123657227, "learning_rate": 8.296749794524094e-05, "loss": 0.0339, "step": 30730 }, { "grad_norm": 0.13992667198181152, "learning_rate": 8.295506475862332e-05, "loss": 0.019, "step": 30740 }, { "grad_norm": 0.4487074911594391, "learning_rate": 8.294262796808933e-05, "loss": 0.0305, "step": 30750 }, { "grad_norm": 0.17886099219322205, "learning_rate": 8.293018757499909e-05, "loss": 0.0181, "step": 30760 }, { "grad_norm": 0.2749488055706024, "learning_rate": 8.291774358071305e-05, "loss": 0.0199, "step": 30770 }, { "grad_norm": 0.4500928819179535, "learning_rate": 8.290529598659205e-05, "loss": 0.0251, "step": 30780 }, { "grad_norm": 0.17472505569458008, "learning_rate": 8.289284479399738e-05, "loss": 0.0294, "step": 30790 }, { "grad_norm": 0.2174430638551712, "learning_rate": 8.288039000429064e-05, "loss": 0.0236, "step": 30800 }, { "grad_norm": 0.2931603789329529, "learning_rate": 8.286793161883388e-05, "loss": 0.0322, "step": 30810 }, { "grad_norm": 0.27287477254867554, "learning_rate": 8.285546963898954e-05, "loss": 0.0296, "step": 30820 }, { "grad_norm": 0.3471953570842743, "learning_rate": 8.284300406612044e-05, "loss": 0.0365, "step": 30830 }, { "grad_norm": 0.36792874336242676, "learning_rate": 8.283053490158978e-05, "loss": 0.0269, "step": 30840 }, { "grad_norm": 0.22192327678203583, "learning_rate": 8.28180621467612e-05, "loss": 0.0198, "step": 30850 }, { "grad_norm": 0.1485622227191925, "learning_rate": 8.280558580299867e-05, "loss": 0.0361, "step": 30860 }, { "grad_norm": 0.2885468602180481, "learning_rate": 8.279310587166661e-05, "loss": 0.0196, "step": 30870 }, { "grad_norm": 0.13289876282215118, "learning_rate": 8.278062235412978e-05, "loss": 0.0281, "step": 30880 }, { "grad_norm": 0.3209797441959381, "learning_rate": 8.276813525175339e-05, "loss": 0.0279, "step": 30890 }, { "grad_norm": 0.4036068618297577, "learning_rate": 8.2755644565903e-05, "loss": 0.0247, "step": 30900 }, { "grad_norm": 0.3834522068500519, "learning_rate": 8.274315029794454e-05, "loss": 0.0305, "step": 30910 }, { "grad_norm": 0.32047051191329956, "learning_rate": 8.273065244924443e-05, "loss": 0.0322, "step": 30920 }, { "grad_norm": 0.2216407209634781, "learning_rate": 8.271815102116936e-05, "loss": 0.0216, "step": 30930 }, { "grad_norm": 0.26801466941833496, "learning_rate": 8.270564601508648e-05, "loss": 0.0214, "step": 30940 }, { "grad_norm": 0.28672394156455994, "learning_rate": 8.269313743236333e-05, "loss": 0.0302, "step": 30950 }, { "grad_norm": 0.3071592152118683, "learning_rate": 8.268062527436783e-05, "loss": 0.0221, "step": 30960 }, { "grad_norm": 0.42565685510635376, "learning_rate": 8.266810954246828e-05, "loss": 0.0311, "step": 30970 }, { "grad_norm": 0.18274199962615967, "learning_rate": 8.265559023803338e-05, "loss": 0.0185, "step": 30980 }, { "grad_norm": 0.4611968696117401, "learning_rate": 8.264306736243223e-05, "loss": 0.0359, "step": 30990 }, { "grad_norm": 0.4238141179084778, "learning_rate": 8.263054091703432e-05, "loss": 0.0276, "step": 31000 }, { "grad_norm": 0.3426406681537628, "learning_rate": 8.26180109032095e-05, "loss": 0.0226, "step": 31010 }, { "grad_norm": 0.23743265867233276, "learning_rate": 8.260547732232807e-05, "loss": 0.0201, "step": 31020 }, { "grad_norm": 0.24405957758426666, "learning_rate": 8.259294017576063e-05, "loss": 0.0272, "step": 31030 }, { "grad_norm": 0.2348044067621231, "learning_rate": 8.258039946487828e-05, "loss": 0.0275, "step": 31040 }, { "grad_norm": 0.316907674074173, "learning_rate": 8.256785519105241e-05, "loss": 0.0201, "step": 31050 }, { "grad_norm": 0.2650400400161743, "learning_rate": 8.255530735565488e-05, "loss": 0.0179, "step": 31060 }, { "grad_norm": 0.20619317889213562, "learning_rate": 8.254275596005787e-05, "loss": 0.0217, "step": 31070 }, { "grad_norm": 0.2984901964664459, "learning_rate": 8.2530201005634e-05, "loss": 0.0239, "step": 31080 }, { "grad_norm": 0.21606586873531342, "learning_rate": 8.251764249375626e-05, "loss": 0.0288, "step": 31090 }, { "grad_norm": 0.1915954351425171, "learning_rate": 8.250508042579803e-05, "loss": 0.0236, "step": 31100 }, { "grad_norm": 0.19545678794384003, "learning_rate": 8.249251480313307e-05, "loss": 0.0264, "step": 31110 }, { "grad_norm": 0.36265870928764343, "learning_rate": 8.247994562713555e-05, "loss": 0.0215, "step": 31120 }, { "grad_norm": 0.523156464099884, "learning_rate": 8.246737289917998e-05, "loss": 0.03, "step": 31130 }, { "grad_norm": 0.20452792942523956, "learning_rate": 8.245479662064135e-05, "loss": 0.0305, "step": 31140 }, { "grad_norm": 0.32227301597595215, "learning_rate": 8.244221679289496e-05, "loss": 0.0264, "step": 31150 }, { "grad_norm": 0.196502223610878, "learning_rate": 8.242963341731652e-05, "loss": 0.0263, "step": 31160 }, { "grad_norm": 0.33398082852363586, "learning_rate": 8.24170464952821e-05, "loss": 0.0294, "step": 31170 }, { "grad_norm": 0.24921727180480957, "learning_rate": 8.240445602816824e-05, "loss": 0.0373, "step": 31180 }, { "grad_norm": 0.2763511836528778, "learning_rate": 8.239186201735179e-05, "loss": 0.0321, "step": 31190 }, { "grad_norm": 0.20732209086418152, "learning_rate": 8.237926446420998e-05, "loss": 0.0254, "step": 31200 }, { "grad_norm": 0.2411726415157318, "learning_rate": 8.236666337012052e-05, "loss": 0.0229, "step": 31210 }, { "grad_norm": 0.22071290016174316, "learning_rate": 8.23540587364614e-05, "loss": 0.0292, "step": 31220 }, { "grad_norm": 0.18982678651809692, "learning_rate": 8.234145056461107e-05, "loss": 0.027, "step": 31230 }, { "grad_norm": 0.3292600214481354, "learning_rate": 8.232883885594832e-05, "loss": 0.0328, "step": 31240 }, { "grad_norm": 0.37821850180625916, "learning_rate": 8.231622361185236e-05, "loss": 0.0335, "step": 31250 }, { "grad_norm": 0.18006306886672974, "learning_rate": 8.230360483370278e-05, "loss": 0.0404, "step": 31260 }, { "grad_norm": 0.19904227554798126, "learning_rate": 8.229098252287953e-05, "loss": 0.015, "step": 31270 }, { "grad_norm": 0.29194289445877075, "learning_rate": 8.2278356680763e-05, "loss": 0.0334, "step": 31280 }, { "grad_norm": 0.3484329283237457, "learning_rate": 8.22657273087339e-05, "loss": 0.0319, "step": 31290 }, { "grad_norm": 0.25688931345939636, "learning_rate": 8.225309440817336e-05, "loss": 0.0318, "step": 31300 }, { "grad_norm": 0.3130180835723877, "learning_rate": 8.224045798046293e-05, "loss": 0.0395, "step": 31310 }, { "grad_norm": 0.3237096965312958, "learning_rate": 8.22278180269845e-05, "loss": 0.0205, "step": 31320 }, { "grad_norm": 0.4375245273113251, "learning_rate": 8.221517454912031e-05, "loss": 0.0228, "step": 31330 }, { "grad_norm": 0.3421066105365753, "learning_rate": 8.220252754825308e-05, "loss": 0.0262, "step": 31340 }, { "grad_norm": 0.31793078780174255, "learning_rate": 8.218987702576586e-05, "loss": 0.0253, "step": 31350 }, { "grad_norm": 0.24115218222141266, "learning_rate": 8.217722298304207e-05, "loss": 0.0408, "step": 31360 }, { "grad_norm": 0.5099972486495972, "learning_rate": 8.216456542146557e-05, "loss": 0.0264, "step": 31370 }, { "grad_norm": 0.23226714134216309, "learning_rate": 8.215190434242055e-05, "loss": 0.0214, "step": 31380 }, { "grad_norm": 0.24196204543113708, "learning_rate": 8.213923974729161e-05, "loss": 0.0281, "step": 31390 }, { "grad_norm": 0.17821525037288666, "learning_rate": 8.212657163746373e-05, "loss": 0.0266, "step": 31400 }, { "grad_norm": 0.20632071793079376, "learning_rate": 8.211390001432227e-05, "loss": 0.0304, "step": 31410 }, { "grad_norm": 0.17002998292446136, "learning_rate": 8.210122487925297e-05, "loss": 0.024, "step": 31420 }, { "grad_norm": 0.38393303751945496, "learning_rate": 8.208854623364202e-05, "loss": 0.0293, "step": 31430 }, { "grad_norm": 0.22480109333992004, "learning_rate": 8.207586407887589e-05, "loss": 0.0211, "step": 31440 }, { "grad_norm": 0.20733143389225006, "learning_rate": 8.206317841634148e-05, "loss": 0.0193, "step": 31450 }, { "grad_norm": 0.1842241734266281, "learning_rate": 8.205048924742609e-05, "loss": 0.029, "step": 31460 }, { "grad_norm": 0.21976880729198456, "learning_rate": 8.203779657351738e-05, "loss": 0.0227, "step": 31470 }, { "grad_norm": 0.24517615139484406, "learning_rate": 8.20251003960034e-05, "loss": 0.0327, "step": 31480 }, { "grad_norm": 0.25390180945396423, "learning_rate": 8.201240071627258e-05, "loss": 0.022, "step": 31490 }, { "grad_norm": 0.18127310276031494, "learning_rate": 8.199969753571377e-05, "loss": 0.0278, "step": 31500 }, { "grad_norm": 0.5122684836387634, "learning_rate": 8.198699085571615e-05, "loss": 0.0285, "step": 31510 }, { "grad_norm": 0.32593652606010437, "learning_rate": 8.197428067766928e-05, "loss": 0.0289, "step": 31520 }, { "grad_norm": 0.3008377254009247, "learning_rate": 8.196156700296316e-05, "loss": 0.031, "step": 31530 }, { "grad_norm": 0.40439608693122864, "learning_rate": 8.194884983298814e-05, "loss": 0.0219, "step": 31540 }, { "grad_norm": 0.37996548414230347, "learning_rate": 8.193612916913491e-05, "loss": 0.0319, "step": 31550 }, { "grad_norm": 0.5674546360969543, "learning_rate": 8.192340501279463e-05, "loss": 0.0403, "step": 31560 }, { "grad_norm": 0.4730892777442932, "learning_rate": 8.191067736535876e-05, "loss": 0.0371, "step": 31570 }, { "grad_norm": 0.2687320113182068, "learning_rate": 8.18979462282192e-05, "loss": 0.0278, "step": 31580 }, { "grad_norm": 0.2962340712547302, "learning_rate": 8.188521160276819e-05, "loss": 0.0265, "step": 31590 }, { "grad_norm": 0.1778809130191803, "learning_rate": 8.187247349039837e-05, "loss": 0.0185, "step": 31600 }, { "grad_norm": 0.3249618113040924, "learning_rate": 8.185973189250278e-05, "loss": 0.0398, "step": 31610 }, { "grad_norm": 0.32795509696006775, "learning_rate": 8.184698681047482e-05, "loss": 0.0266, "step": 31620 }, { "grad_norm": 0.17014147341251373, "learning_rate": 8.183423824570827e-05, "loss": 0.0219, "step": 31630 }, { "grad_norm": 0.3958584666252136, "learning_rate": 8.182148619959725e-05, "loss": 0.0354, "step": 31640 }, { "grad_norm": 0.3821175694465637, "learning_rate": 8.180873067353636e-05, "loss": 0.0319, "step": 31650 }, { "grad_norm": 0.20628058910369873, "learning_rate": 8.179597166892052e-05, "loss": 0.0221, "step": 31660 }, { "grad_norm": 0.32990896701812744, "learning_rate": 8.178320918714501e-05, "loss": 0.0277, "step": 31670 }, { "grad_norm": 0.2581448554992676, "learning_rate": 8.177044322960554e-05, "loss": 0.0265, "step": 31680 }, { "grad_norm": 0.18167026340961456, "learning_rate": 8.175767379769816e-05, "loss": 0.0215, "step": 31690 }, { "grad_norm": 0.20542800426483154, "learning_rate": 8.174490089281932e-05, "loss": 0.0255, "step": 31700 }, { "grad_norm": 0.33688005805015564, "learning_rate": 8.173212451636584e-05, "loss": 0.0191, "step": 31710 }, { "grad_norm": 0.22522176802158356, "learning_rate": 8.171934466973493e-05, "loss": 0.023, "step": 31720 }, { "grad_norm": 0.34164395928382874, "learning_rate": 8.170656135432418e-05, "loss": 0.0245, "step": 31730 }, { "grad_norm": 0.19672611355781555, "learning_rate": 8.169377457153155e-05, "loss": 0.0366, "step": 31740 }, { "grad_norm": 0.34905144572257996, "learning_rate": 8.168098432275539e-05, "loss": 0.023, "step": 31750 }, { "grad_norm": 0.44510143995285034, "learning_rate": 8.166819060939442e-05, "loss": 0.0287, "step": 31760 }, { "grad_norm": 0.3672085106372833, "learning_rate": 8.165539343284772e-05, "loss": 0.0311, "step": 31770 }, { "grad_norm": 0.4325260519981384, "learning_rate": 8.16425927945148e-05, "loss": 0.0444, "step": 31780 }, { "grad_norm": 0.1880146563053131, "learning_rate": 8.162978869579551e-05, "loss": 0.0456, "step": 31790 }, { "grad_norm": 0.23484133183956146, "learning_rate": 8.161698113809007e-05, "loss": 0.0242, "step": 31800 }, { "grad_norm": 0.37391215562820435, "learning_rate": 8.160417012279911e-05, "loss": 0.0348, "step": 31810 }, { "grad_norm": 0.22826313972473145, "learning_rate": 8.159135565132363e-05, "loss": 0.0202, "step": 31820 }, { "grad_norm": 0.29927128553390503, "learning_rate": 8.157853772506498e-05, "loss": 0.0325, "step": 31830 }, { "grad_norm": 0.36649566888809204, "learning_rate": 8.156571634542494e-05, "loss": 0.0245, "step": 31840 }, { "grad_norm": 0.28139954805374146, "learning_rate": 8.15528915138056e-05, "loss": 0.0379, "step": 31850 }, { "grad_norm": 0.3843802213668823, "learning_rate": 8.154006323160949e-05, "loss": 0.0259, "step": 31860 }, { "grad_norm": 0.20765161514282227, "learning_rate": 8.152723150023949e-05, "loss": 0.0304, "step": 31870 }, { "grad_norm": 0.2970271408557892, "learning_rate": 8.151439632109886e-05, "loss": 0.0343, "step": 31880 }, { "grad_norm": 0.12280984222888947, "learning_rate": 8.150155769559122e-05, "loss": 0.0369, "step": 31890 }, { "grad_norm": 0.3197631537914276, "learning_rate": 8.148871562512058e-05, "loss": 0.0255, "step": 31900 }, { "grad_norm": 0.2763584852218628, "learning_rate": 8.147587011109136e-05, "loss": 0.0291, "step": 31910 }, { "grad_norm": 0.40391889214515686, "learning_rate": 8.14630211549083e-05, "loss": 0.0517, "step": 31920 }, { "grad_norm": 0.24233238399028778, "learning_rate": 8.145016875797655e-05, "loss": 0.022, "step": 31930 }, { "grad_norm": 0.3059731125831604, "learning_rate": 8.143731292170164e-05, "loss": 0.0271, "step": 31940 }, { "grad_norm": 0.3888488709926605, "learning_rate": 8.142445364748944e-05, "loss": 0.0368, "step": 31950 }, { "grad_norm": 0.40900933742523193, "learning_rate": 8.141159093674624e-05, "loss": 0.0328, "step": 31960 }, { "grad_norm": 0.21434570848941803, "learning_rate": 8.139872479087869e-05, "loss": 0.0265, "step": 31970 }, { "grad_norm": 0.2915875017642975, "learning_rate": 8.13858552112938e-05, "loss": 0.0227, "step": 31980 }, { "grad_norm": 0.22283823788166046, "learning_rate": 8.137298219939895e-05, "loss": 0.0281, "step": 31990 }, { "grad_norm": 0.35005027055740356, "learning_rate": 8.136010575660196e-05, "loss": 0.0315, "step": 32000 }, { "grad_norm": 0.29936903715133667, "learning_rate": 8.134722588431095e-05, "loss": 0.0239, "step": 32010 }, { "grad_norm": 0.3230767250061035, "learning_rate": 8.133434258393444e-05, "loss": 0.0285, "step": 32020 }, { "grad_norm": 0.15867683291435242, "learning_rate": 8.132145585688134e-05, "loss": 0.0175, "step": 32030 }, { "grad_norm": 0.36984163522720337, "learning_rate": 8.130856570456093e-05, "loss": 0.0279, "step": 32040 }, { "grad_norm": 0.5035372972488403, "learning_rate": 8.129567212838283e-05, "loss": 0.0226, "step": 32050 }, { "grad_norm": 0.35925331711769104, "learning_rate": 8.128277512975708e-05, "loss": 0.032, "step": 32060 }, { "grad_norm": 0.24372349679470062, "learning_rate": 8.126987471009408e-05, "loss": 0.0285, "step": 32070 }, { "grad_norm": 0.37705036997795105, "learning_rate": 8.125697087080459e-05, "loss": 0.0198, "step": 32080 }, { "grad_norm": 0.238324373960495, "learning_rate": 8.124406361329976e-05, "loss": 0.028, "step": 32090 }, { "grad_norm": 0.27566957473754883, "learning_rate": 8.123115293899111e-05, "loss": 0.0294, "step": 32100 }, { "grad_norm": 0.14985525608062744, "learning_rate": 8.121823884929055e-05, "loss": 0.0193, "step": 32110 }, { "grad_norm": 0.09651214629411697, "learning_rate": 8.12053213456103e-05, "loss": 0.0255, "step": 32120 }, { "grad_norm": 0.2778621315956116, "learning_rate": 8.119240042936303e-05, "loss": 0.0341, "step": 32130 }, { "grad_norm": 0.24613530933856964, "learning_rate": 8.117947610196175e-05, "loss": 0.0247, "step": 32140 }, { "grad_norm": 0.5851697325706482, "learning_rate": 8.116654836481982e-05, "loss": 0.0267, "step": 32150 }, { "grad_norm": 0.2364228218793869, "learning_rate": 8.115361721935106e-05, "loss": 0.0252, "step": 32160 }, { "grad_norm": 0.1861306130886078, "learning_rate": 8.114068266696953e-05, "loss": 0.0239, "step": 32170 }, { "grad_norm": 0.3751550316810608, "learning_rate": 8.112774470908978e-05, "loss": 0.0212, "step": 32180 }, { "grad_norm": 0.2452053725719452, "learning_rate": 8.111480334712665e-05, "loss": 0.0156, "step": 32190 }, { "grad_norm": 0.10245440155267715, "learning_rate": 8.110185858249542e-05, "loss": 0.0196, "step": 32200 }, { "grad_norm": 0.1947399079799652, "learning_rate": 8.108891041661168e-05, "loss": 0.0301, "step": 32210 }, { "grad_norm": 0.2520877718925476, "learning_rate": 8.107595885089146e-05, "loss": 0.0299, "step": 32220 }, { "grad_norm": 0.25001060962677, "learning_rate": 8.106300388675108e-05, "loss": 0.0211, "step": 32230 }, { "grad_norm": 0.3721005320549011, "learning_rate": 8.105004552560731e-05, "loss": 0.0416, "step": 32240 }, { "grad_norm": 0.25625818967819214, "learning_rate": 8.103708376887724e-05, "loss": 0.0335, "step": 32250 }, { "grad_norm": 0.23903432488441467, "learning_rate": 8.102411861797836e-05, "loss": 0.0277, "step": 32260 }, { "grad_norm": 0.20665408670902252, "learning_rate": 8.101115007432851e-05, "loss": 0.0195, "step": 32270 }, { "grad_norm": 0.199667826294899, "learning_rate": 8.09981781393459e-05, "loss": 0.0197, "step": 32280 }, { "grad_norm": 0.19655002653598785, "learning_rate": 8.098520281444915e-05, "loss": 0.0204, "step": 32290 }, { "grad_norm": 0.40040475130081177, "learning_rate": 8.09722241010572e-05, "loss": 0.0234, "step": 32300 }, { "grad_norm": 0.12906353175640106, "learning_rate": 8.095924200058939e-05, "loss": 0.0217, "step": 32310 }, { "grad_norm": 0.31959637999534607, "learning_rate": 8.094625651446541e-05, "loss": 0.0121, "step": 32320 }, { "grad_norm": 0.24482868611812592, "learning_rate": 8.093326764410536e-05, "loss": 0.0228, "step": 32330 }, { "grad_norm": 0.23125137388706207, "learning_rate": 8.092027539092966e-05, "loss": 0.0251, "step": 32340 }, { "grad_norm": 0.24064728617668152, "learning_rate": 8.090727975635913e-05, "loss": 0.0209, "step": 32350 }, { "grad_norm": 0.2988843321800232, "learning_rate": 8.089428074181497e-05, "loss": 0.0263, "step": 32360 }, { "grad_norm": 0.43962982296943665, "learning_rate": 8.088127834871871e-05, "loss": 0.0211, "step": 32370 }, { "grad_norm": 0.2363724410533905, "learning_rate": 8.086827257849226e-05, "loss": 0.0214, "step": 32380 }, { "grad_norm": 0.17912490665912628, "learning_rate": 8.085526343255795e-05, "loss": 0.0162, "step": 32390 }, { "grad_norm": 0.31628093123435974, "learning_rate": 8.084225091233842e-05, "loss": 0.0166, "step": 32400 }, { "grad_norm": 0.30814358592033386, "learning_rate": 8.082923501925668e-05, "loss": 0.0215, "step": 32410 }, { "grad_norm": 0.1594526320695877, "learning_rate": 8.081621575473617e-05, "loss": 0.0261, "step": 32420 }, { "grad_norm": 0.18643330037593842, "learning_rate": 8.080319312020064e-05, "loss": 0.027, "step": 32430 }, { "grad_norm": 0.20380181074142456, "learning_rate": 8.079016711707421e-05, "loss": 0.0148, "step": 32440 }, { "grad_norm": 0.16522274911403656, "learning_rate": 8.077713774678139e-05, "loss": 0.0198, "step": 32450 }, { "grad_norm": 0.1380888819694519, "learning_rate": 8.076410501074707e-05, "loss": 0.0187, "step": 32460 }, { "grad_norm": 0.20204545557498932, "learning_rate": 8.075106891039647e-05, "loss": 0.0315, "step": 32470 }, { "grad_norm": 0.19165116548538208, "learning_rate": 8.073802944715523e-05, "loss": 0.0316, "step": 32480 }, { "grad_norm": 0.33707931637763977, "learning_rate": 8.072498662244929e-05, "loss": 0.0227, "step": 32490 }, { "grad_norm": 0.2364356517791748, "learning_rate": 8.0711940437705e-05, "loss": 0.0252, "step": 32500 }, { "grad_norm": 0.12729401886463165, "learning_rate": 8.06988908943491e-05, "loss": 0.0194, "step": 32510 }, { "grad_norm": 0.18280288577079773, "learning_rate": 8.068583799380863e-05, "loss": 0.0224, "step": 32520 }, { "grad_norm": 0.20061449706554413, "learning_rate": 8.067278173751104e-05, "loss": 0.0189, "step": 32530 }, { "grad_norm": 0.33282822370529175, "learning_rate": 8.065972212688417e-05, "loss": 0.0213, "step": 32540 }, { "grad_norm": 0.7883819937705994, "learning_rate": 8.064665916335618e-05, "loss": 0.0309, "step": 32550 }, { "grad_norm": 0.21970221400260925, "learning_rate": 8.063359284835564e-05, "loss": 0.024, "step": 32560 }, { "grad_norm": 0.44340577721595764, "learning_rate": 8.062052318331142e-05, "loss": 0.0282, "step": 32570 }, { "grad_norm": 0.4229460060596466, "learning_rate": 8.060745016965283e-05, "loss": 0.0238, "step": 32580 }, { "grad_norm": 0.25735369324684143, "learning_rate": 8.059437380880952e-05, "loss": 0.0263, "step": 32590 }, { "grad_norm": 0.3545056879520416, "learning_rate": 8.058129410221146e-05, "loss": 0.0232, "step": 32600 }, { "grad_norm": 0.18884025514125824, "learning_rate": 8.056821105128908e-05, "loss": 0.0241, "step": 32610 }, { "grad_norm": 0.13680213689804077, "learning_rate": 8.05551246574731e-05, "loss": 0.0275, "step": 32620 }, { "grad_norm": 0.10759256780147552, "learning_rate": 8.05420349221946e-05, "loss": 0.023, "step": 32630 }, { "grad_norm": 0.23965013027191162, "learning_rate": 8.05289418468851e-05, "loss": 0.0212, "step": 32640 }, { "grad_norm": 0.37610775232315063, "learning_rate": 8.051584543297642e-05, "loss": 0.0296, "step": 32650 }, { "grad_norm": 0.2980043292045593, "learning_rate": 8.050274568190074e-05, "loss": 0.0223, "step": 32660 }, { "grad_norm": 0.16584594547748566, "learning_rate": 8.048964259509067e-05, "loss": 0.0285, "step": 32670 }, { "grad_norm": 0.3948187530040741, "learning_rate": 8.047653617397914e-05, "loss": 0.024, "step": 32680 }, { "grad_norm": 0.36606746912002563, "learning_rate": 8.046342641999941e-05, "loss": 0.0233, "step": 32690 }, { "grad_norm": 0.3204219341278076, "learning_rate": 8.045031333458517e-05, "loss": 0.0267, "step": 32700 }, { "grad_norm": 0.40506163239479065, "learning_rate": 8.043719691917047e-05, "loss": 0.0293, "step": 32710 }, { "grad_norm": 0.4468883275985718, "learning_rate": 8.042407717518966e-05, "loss": 0.0248, "step": 32720 }, { "grad_norm": 0.2805582582950592, "learning_rate": 8.041095410407751e-05, "loss": 0.0344, "step": 32730 }, { "grad_norm": 0.1756720393896103, "learning_rate": 8.039782770726913e-05, "loss": 0.0257, "step": 32740 }, { "grad_norm": 0.31883201003074646, "learning_rate": 8.038469798620004e-05, "loss": 0.0245, "step": 32750 }, { "grad_norm": 0.5051984786987305, "learning_rate": 8.037156494230604e-05, "loss": 0.0275, "step": 32760 }, { "grad_norm": 0.2123241275548935, "learning_rate": 8.035842857702338e-05, "loss": 0.0271, "step": 32770 }, { "grad_norm": 0.3465605080127716, "learning_rate": 8.03452888917886e-05, "loss": 0.0322, "step": 32780 }, { "grad_norm": 0.14785555005073547, "learning_rate": 8.033214588803866e-05, "loss": 0.0153, "step": 32790 }, { "grad_norm": 0.18928605318069458, "learning_rate": 8.031899956721083e-05, "loss": 0.0229, "step": 32800 }, { "grad_norm": 0.21635259687900543, "learning_rate": 8.030584993074282e-05, "loss": 0.0188, "step": 32810 }, { "grad_norm": 0.21834218502044678, "learning_rate": 8.02926969800726e-05, "loss": 0.0272, "step": 32820 }, { "grad_norm": 0.17135776579380035, "learning_rate": 8.027954071663859e-05, "loss": 0.0309, "step": 32830 }, { "grad_norm": 0.1804855465888977, "learning_rate": 8.026638114187954e-05, "loss": 0.0184, "step": 32840 }, { "grad_norm": 0.32630500197410583, "learning_rate": 8.025321825723456e-05, "loss": 0.0262, "step": 32850 }, { "grad_norm": 0.32663464546203613, "learning_rate": 8.02400520641431e-05, "loss": 0.0353, "step": 32860 }, { "grad_norm": 0.2153572142124176, "learning_rate": 8.022688256404501e-05, "loss": 0.0213, "step": 32870 }, { "grad_norm": 0.1546303778886795, "learning_rate": 8.02137097583805e-05, "loss": 0.0211, "step": 32880 }, { "grad_norm": 0.279723584651947, "learning_rate": 8.02005336485901e-05, "loss": 0.0301, "step": 32890 }, { "grad_norm": 0.21567568182945251, "learning_rate": 8.018735423611476e-05, "loss": 0.0348, "step": 32900 }, { "grad_norm": 0.40199610590934753, "learning_rate": 8.017417152239574e-05, "loss": 0.0323, "step": 32910 }, { "grad_norm": 0.21769453585147858, "learning_rate": 8.01609855088747e-05, "loss": 0.0278, "step": 32920 }, { "grad_norm": 0.2217356115579605, "learning_rate": 8.014779619699362e-05, "loss": 0.0202, "step": 32930 }, { "grad_norm": 0.21317902207374573, "learning_rate": 8.013460358819489e-05, "loss": 0.023, "step": 32940 }, { "grad_norm": 0.3814140856266022, "learning_rate": 8.01214076839212e-05, "loss": 0.0355, "step": 32950 }, { "grad_norm": 0.29040688276290894, "learning_rate": 8.010820848561565e-05, "loss": 0.0257, "step": 32960 }, { "grad_norm": 0.34748443961143494, "learning_rate": 8.009500599472171e-05, "loss": 0.0254, "step": 32970 }, { "grad_norm": 0.19471915066242218, "learning_rate": 8.008180021268314e-05, "loss": 0.0253, "step": 32980 }, { "grad_norm": 0.2282225638628006, "learning_rate": 8.006859114094414e-05, "loss": 0.0231, "step": 32990 }, { "grad_norm": 0.1481754034757614, "learning_rate": 8.005537878094921e-05, "loss": 0.0224, "step": 33000 }, { "grad_norm": 0.32125258445739746, "learning_rate": 8.004216313414323e-05, "loss": 0.0432, "step": 33010 }, { "grad_norm": 0.16978046298027039, "learning_rate": 8.002894420197149e-05, "loss": 0.026, "step": 33020 }, { "grad_norm": 0.24025094509124756, "learning_rate": 8.001572198587954e-05, "loss": 0.0301, "step": 33030 }, { "grad_norm": 0.21489548683166504, "learning_rate": 8.000249648731338e-05, "loss": 0.0242, "step": 33040 }, { "grad_norm": 0.6121675372123718, "learning_rate": 7.998926770771928e-05, "loss": 0.0476, "step": 33050 }, { "grad_norm": 0.2196202278137207, "learning_rate": 7.997603564854397e-05, "loss": 0.0218, "step": 33060 }, { "grad_norm": 0.34840184450149536, "learning_rate": 7.996280031123448e-05, "loss": 0.0374, "step": 33070 }, { "grad_norm": 0.2356550693511963, "learning_rate": 7.994956169723818e-05, "loss": 0.044, "step": 33080 }, { "grad_norm": 0.37034717202186584, "learning_rate": 7.993631980800285e-05, "loss": 0.0282, "step": 33090 }, { "grad_norm": 0.12813381850719452, "learning_rate": 7.992307464497659e-05, "loss": 0.0349, "step": 33100 }, { "grad_norm": 0.2071453034877777, "learning_rate": 7.990982620960787e-05, "loss": 0.0222, "step": 33110 }, { "grad_norm": 0.2605319917201996, "learning_rate": 7.989657450334554e-05, "loss": 0.0277, "step": 33120 }, { "grad_norm": 0.18957626819610596, "learning_rate": 7.988331952763877e-05, "loss": 0.0256, "step": 33130 }, { "grad_norm": 0.3217001259326935, "learning_rate": 7.987006128393709e-05, "loss": 0.0362, "step": 33140 }, { "grad_norm": 0.274166464805603, "learning_rate": 7.985679977369043e-05, "loss": 0.0202, "step": 33150 }, { "grad_norm": 0.34407320618629456, "learning_rate": 7.984353499834902e-05, "loss": 0.0296, "step": 33160 }, { "grad_norm": 0.4097974896430969, "learning_rate": 7.983026695936351e-05, "loss": 0.0266, "step": 33170 }, { "grad_norm": 0.29810017347335815, "learning_rate": 7.981699565818486e-05, "loss": 0.0214, "step": 33180 }, { "grad_norm": 0.29043588042259216, "learning_rate": 7.980372109626437e-05, "loss": 0.0237, "step": 33190 }, { "grad_norm": 0.2995639145374298, "learning_rate": 7.979044327505375e-05, "loss": 0.0238, "step": 33200 }, { "grad_norm": 0.20912471413612366, "learning_rate": 7.977716219600506e-05, "loss": 0.0278, "step": 33210 }, { "grad_norm": 0.3058413863182068, "learning_rate": 7.97638778605707e-05, "loss": 0.0208, "step": 33220 }, { "grad_norm": 0.137546107172966, "learning_rate": 7.975059027020338e-05, "loss": 0.0386, "step": 33230 }, { "grad_norm": 0.10962239652872086, "learning_rate": 7.973729942635623e-05, "loss": 0.0331, "step": 33240 }, { "grad_norm": 0.3126739263534546, "learning_rate": 7.972400533048273e-05, "loss": 0.0283, "step": 33250 }, { "grad_norm": 0.20664075016975403, "learning_rate": 7.97107079840367e-05, "loss": 0.0228, "step": 33260 }, { "grad_norm": 0.13828571140766144, "learning_rate": 7.969740738847231e-05, "loss": 0.022, "step": 33270 }, { "grad_norm": 0.260721355676651, "learning_rate": 7.968410354524411e-05, "loss": 0.0334, "step": 33280 }, { "grad_norm": 0.23221376538276672, "learning_rate": 7.967079645580697e-05, "loss": 0.0311, "step": 33290 }, { "grad_norm": 0.3331027925014496, "learning_rate": 7.965748612161612e-05, "loss": 0.0282, "step": 33300 }, { "grad_norm": 0.20547981560230255, "learning_rate": 7.96441725441272e-05, "loss": 0.027, "step": 33310 }, { "grad_norm": 0.11204421520233154, "learning_rate": 7.963085572479614e-05, "loss": 0.025, "step": 33320 }, { "grad_norm": 0.1790313422679901, "learning_rate": 7.961753566507924e-05, "loss": 0.0207, "step": 33330 }, { "grad_norm": 0.3589850962162018, "learning_rate": 7.960421236643316e-05, "loss": 0.0254, "step": 33340 }, { "grad_norm": 0.26547181606292725, "learning_rate": 7.959088583031496e-05, "loss": 0.023, "step": 33350 }, { "grad_norm": 0.12237624078989029, "learning_rate": 7.957755605818194e-05, "loss": 0.0167, "step": 33360 }, { "grad_norm": 0.3244917392730713, "learning_rate": 7.956422305149185e-05, "loss": 0.0267, "step": 33370 }, { "grad_norm": 0.3572528660297394, "learning_rate": 7.95508868117028e-05, "loss": 0.027, "step": 33380 }, { "grad_norm": 0.49403324723243713, "learning_rate": 7.953754734027318e-05, "loss": 0.0274, "step": 33390 }, { "grad_norm": 0.1471443474292755, "learning_rate": 7.952420463866182e-05, "loss": 0.0157, "step": 33400 }, { "grad_norm": 0.13659512996673584, "learning_rate": 7.951085870832782e-05, "loss": 0.0199, "step": 33410 }, { "grad_norm": 0.2694006860256195, "learning_rate": 7.949750955073067e-05, "loss": 0.0211, "step": 33420 }, { "grad_norm": 0.2326306253671646, "learning_rate": 7.948415716733022e-05, "loss": 0.0341, "step": 33430 }, { "grad_norm": 0.22196152806282043, "learning_rate": 7.947080155958669e-05, "loss": 0.033, "step": 33440 }, { "grad_norm": 0.21645577251911163, "learning_rate": 7.94574427289606e-05, "loss": 0.0214, "step": 33450 }, { "grad_norm": 0.3069418668746948, "learning_rate": 7.944408067691284e-05, "loss": 0.025, "step": 33460 }, { "grad_norm": 0.298202782869339, "learning_rate": 7.943071540490473e-05, "loss": 0.0216, "step": 33470 }, { "grad_norm": 0.22350165247917175, "learning_rate": 7.94173469143978e-05, "loss": 0.0196, "step": 33480 }, { "grad_norm": 0.28526902198791504, "learning_rate": 7.940397520685406e-05, "loss": 0.023, "step": 33490 }, { "grad_norm": 0.3362792432308197, "learning_rate": 7.939060028373577e-05, "loss": 0.0243, "step": 33500 }, { "grad_norm": 0.17462241649627686, "learning_rate": 7.937722214650565e-05, "loss": 0.0308, "step": 33510 }, { "grad_norm": 0.2091040164232254, "learning_rate": 7.936384079662666e-05, "loss": 0.026, "step": 33520 }, { "grad_norm": 0.3120039403438568, "learning_rate": 7.93504562355622e-05, "loss": 0.0267, "step": 33530 }, { "grad_norm": 0.3159567713737488, "learning_rate": 7.933706846477599e-05, "loss": 0.0273, "step": 33540 }, { "grad_norm": 0.3124980330467224, "learning_rate": 7.932367748573206e-05, "loss": 0.0327, "step": 33550 }, { "grad_norm": 0.2086445838212967, "learning_rate": 7.931028329989485e-05, "loss": 0.0223, "step": 33560 }, { "grad_norm": 0.1992645412683487, "learning_rate": 7.929688590872913e-05, "loss": 0.0371, "step": 33570 }, { "grad_norm": 0.11105220764875412, "learning_rate": 7.928348531370003e-05, "loss": 0.0229, "step": 33580 }, { "grad_norm": 0.3706071972846985, "learning_rate": 7.927008151627297e-05, "loss": 0.0307, "step": 33590 }, { "grad_norm": 0.3578554689884186, "learning_rate": 7.925667451791383e-05, "loss": 0.0207, "step": 33600 }, { "grad_norm": 0.1542283296585083, "learning_rate": 7.924326432008874e-05, "loss": 0.0245, "step": 33610 }, { "grad_norm": 0.25817587971687317, "learning_rate": 7.922985092426422e-05, "loss": 0.0238, "step": 33620 }, { "grad_norm": 0.5146726965904236, "learning_rate": 7.921643433190717e-05, "loss": 0.0296, "step": 33630 }, { "grad_norm": 0.44108423590660095, "learning_rate": 7.920301454448478e-05, "loss": 0.0327, "step": 33640 }, { "grad_norm": 0.2682953178882599, "learning_rate": 7.918959156346461e-05, "loss": 0.02, "step": 33650 }, { "grad_norm": 0.19246801733970642, "learning_rate": 7.91761653903146e-05, "loss": 0.0241, "step": 33660 }, { "grad_norm": 0.29994457960128784, "learning_rate": 7.916273602650302e-05, "loss": 0.026, "step": 33670 }, { "grad_norm": 0.1527402102947235, "learning_rate": 7.914930347349847e-05, "loss": 0.0189, "step": 33680 }, { "grad_norm": 0.16145193576812744, "learning_rate": 7.913586773276992e-05, "loss": 0.0239, "step": 33690 }, { "grad_norm": 0.5953474640846252, "learning_rate": 7.912242880578667e-05, "loss": 0.0248, "step": 33700 }, { "grad_norm": 0.3726680278778076, "learning_rate": 7.910898669401839e-05, "loss": 0.0263, "step": 33710 }, { "grad_norm": 0.47623196244239807, "learning_rate": 7.909554139893511e-05, "loss": 0.0307, "step": 33720 }, { "grad_norm": 0.2386624813079834, "learning_rate": 7.908209292200715e-05, "loss": 0.0172, "step": 33730 }, { "grad_norm": 0.20726965367794037, "learning_rate": 7.906864126470523e-05, "loss": 0.0234, "step": 33740 }, { "grad_norm": 0.33219441771507263, "learning_rate": 7.905518642850041e-05, "loss": 0.0258, "step": 33750 }, { "grad_norm": 0.4078533947467804, "learning_rate": 7.904172841486409e-05, "loss": 0.0233, "step": 33760 }, { "grad_norm": 0.12899774312973022, "learning_rate": 7.902826722526801e-05, "loss": 0.0282, "step": 33770 }, { "grad_norm": 0.23788292706012726, "learning_rate": 7.901480286118427e-05, "loss": 0.0163, "step": 33780 }, { "grad_norm": 0.33170583844184875, "learning_rate": 7.900133532408531e-05, "loss": 0.0236, "step": 33790 }, { "grad_norm": 0.30132925510406494, "learning_rate": 7.898786461544395e-05, "loss": 0.0293, "step": 33800 }, { "grad_norm": 0.3263271152973175, "learning_rate": 7.897439073673325e-05, "loss": 0.0229, "step": 33810 }, { "grad_norm": 0.19706834852695465, "learning_rate": 7.896091368942677e-05, "loss": 0.0282, "step": 33820 }, { "grad_norm": 0.2972095012664795, "learning_rate": 7.894743347499832e-05, "loss": 0.0225, "step": 33830 }, { "grad_norm": 0.3764933943748474, "learning_rate": 7.893395009492203e-05, "loss": 0.0302, "step": 33840 }, { "grad_norm": 0.30097854137420654, "learning_rate": 7.892046355067248e-05, "loss": 0.0332, "step": 33850 }, { "grad_norm": 0.43383166193962097, "learning_rate": 7.890697384372451e-05, "loss": 0.0277, "step": 33860 }, { "grad_norm": 0.5018990635871887, "learning_rate": 7.889348097555336e-05, "loss": 0.0212, "step": 33870 }, { "grad_norm": 0.21497413516044617, "learning_rate": 7.887998494763455e-05, "loss": 0.0248, "step": 33880 }, { "grad_norm": 0.3010425567626953, "learning_rate": 7.886648576144404e-05, "loss": 0.0304, "step": 33890 }, { "grad_norm": 0.35272282361984253, "learning_rate": 7.885298341845802e-05, "loss": 0.0265, "step": 33900 }, { "grad_norm": 0.19816595315933228, "learning_rate": 7.883947792015311e-05, "loss": 0.0264, "step": 33910 }, { "grad_norm": 0.17804177105426788, "learning_rate": 7.882596926800628e-05, "loss": 0.0499, "step": 33920 }, { "grad_norm": 0.16644287109375, "learning_rate": 7.881245746349477e-05, "loss": 0.0243, "step": 33930 }, { "grad_norm": 0.21507270634174347, "learning_rate": 7.879894250809623e-05, "loss": 0.0179, "step": 33940 }, { "grad_norm": 0.14087986946105957, "learning_rate": 7.878542440328865e-05, "loss": 0.0352, "step": 33950 }, { "grad_norm": 0.1519595831632614, "learning_rate": 7.877190315055031e-05, "loss": 0.0195, "step": 33960 }, { "grad_norm": 0.3943532705307007, "learning_rate": 7.875837875135991e-05, "loss": 0.0245, "step": 33970 }, { "grad_norm": 0.26490479707717896, "learning_rate": 7.874485120719646e-05, "loss": 0.0222, "step": 33980 }, { "grad_norm": 0.3307507038116455, "learning_rate": 7.873132051953928e-05, "loss": 0.0211, "step": 33990 }, { "grad_norm": 0.3451266586780548, "learning_rate": 7.87177866898681e-05, "loss": 0.0341, "step": 34000 }, { "grad_norm": 0.20528961718082428, "learning_rate": 7.870424971966294e-05, "loss": 0.0199, "step": 34010 }, { "grad_norm": 0.293555349111557, "learning_rate": 7.869070961040419e-05, "loss": 0.0231, "step": 34020 }, { "grad_norm": 0.22048765420913696, "learning_rate": 7.867716636357257e-05, "loss": 0.0269, "step": 34030 }, { "grad_norm": 0.3167141377925873, "learning_rate": 7.866361998064915e-05, "loss": 0.0322, "step": 34040 }, { "grad_norm": 0.1312921941280365, "learning_rate": 7.865007046311534e-05, "loss": 0.0268, "step": 34050 }, { "grad_norm": 0.45861566066741943, "learning_rate": 7.86365178124529e-05, "loss": 0.03, "step": 34060 }, { "grad_norm": 0.27972400188446045, "learning_rate": 7.862296203014394e-05, "loss": 0.0232, "step": 34070 }, { "grad_norm": 0.34633034467697144, "learning_rate": 7.860940311767088e-05, "loss": 0.032, "step": 34080 }, { "grad_norm": 0.5023788213729858, "learning_rate": 7.85958410765165e-05, "loss": 0.0245, "step": 34090 }, { "grad_norm": 0.13430394232273102, "learning_rate": 7.858227590816394e-05, "loss": 0.0171, "step": 34100 }, { "grad_norm": 0.21799829602241516, "learning_rate": 7.856870761409664e-05, "loss": 0.025, "step": 34110 }, { "grad_norm": 0.2325173318386078, "learning_rate": 7.855513619579846e-05, "loss": 0.0246, "step": 34120 }, { "grad_norm": 0.16378498077392578, "learning_rate": 7.85415616547535e-05, "loss": 0.0276, "step": 34130 }, { "grad_norm": 0.21161295473575592, "learning_rate": 7.852798399244627e-05, "loss": 0.0183, "step": 34140 }, { "grad_norm": 0.23989608883857727, "learning_rate": 7.851440321036161e-05, "loss": 0.0211, "step": 34150 }, { "grad_norm": 0.46752363443374634, "learning_rate": 7.850081930998468e-05, "loss": 0.0338, "step": 34160 }, { "grad_norm": 0.21357856690883636, "learning_rate": 7.8487232292801e-05, "loss": 0.0334, "step": 34170 }, { "grad_norm": 0.20952318608760834, "learning_rate": 7.847364216029642e-05, "loss": 0.0237, "step": 34180 }, { "grad_norm": 0.32068780064582825, "learning_rate": 7.846004891395716e-05, "loss": 0.0242, "step": 34190 }, { "grad_norm": 0.2965408265590668, "learning_rate": 7.844645255526972e-05, "loss": 0.0223, "step": 34200 }, { "grad_norm": 0.2558761239051819, "learning_rate": 7.843285308572101e-05, "loss": 0.0218, "step": 34210 }, { "grad_norm": 0.16315937042236328, "learning_rate": 7.841925050679823e-05, "loss": 0.0202, "step": 34220 }, { "grad_norm": 0.21663957834243774, "learning_rate": 7.840564481998895e-05, "loss": 0.0245, "step": 34230 }, { "grad_norm": 0.32205167412757874, "learning_rate": 7.839203602678105e-05, "loss": 0.0185, "step": 34240 }, { "grad_norm": 0.4337327480316162, "learning_rate": 7.837842412866279e-05, "loss": 0.023, "step": 34250 }, { "grad_norm": 0.2531239986419678, "learning_rate": 7.836480912712272e-05, "loss": 0.031, "step": 34260 }, { "grad_norm": 0.5478608012199402, "learning_rate": 7.835119102364976e-05, "loss": 0.0279, "step": 34270 }, { "grad_norm": 0.21382023394107819, "learning_rate": 7.83375698197332e-05, "loss": 0.0241, "step": 34280 }, { "grad_norm": 0.417947918176651, "learning_rate": 7.83239455168626e-05, "loss": 0.0243, "step": 34290 }, { "grad_norm": 0.2367256134748459, "learning_rate": 7.83103181165279e-05, "loss": 0.02, "step": 34300 }, { "grad_norm": 0.37596625089645386, "learning_rate": 7.829668762021937e-05, "loss": 0.0237, "step": 34310 }, { "grad_norm": 0.3867001533508301, "learning_rate": 7.828305402942764e-05, "loss": 0.0249, "step": 34320 }, { "grad_norm": 0.7256588339805603, "learning_rate": 7.826941734564363e-05, "loss": 0.0209, "step": 34330 }, { "grad_norm": 0.28304189443588257, "learning_rate": 7.825577757035865e-05, "loss": 0.0138, "step": 34340 }, { "grad_norm": 0.3180120587348938, "learning_rate": 7.824213470506431e-05, "loss": 0.0271, "step": 34350 }, { "grad_norm": 0.3282336890697479, "learning_rate": 7.822848875125257e-05, "loss": 0.021, "step": 34360 }, { "grad_norm": 0.3492896258831024, "learning_rate": 7.821483971041576e-05, "loss": 0.0296, "step": 34370 }, { "grad_norm": 0.40386271476745605, "learning_rate": 7.820118758404649e-05, "loss": 0.0278, "step": 34380 }, { "grad_norm": 0.36197736859321594, "learning_rate": 7.818753237363776e-05, "loss": 0.0248, "step": 34390 }, { "grad_norm": 0.20998254418373108, "learning_rate": 7.817387408068286e-05, "loss": 0.0225, "step": 34400 }, { "grad_norm": 0.3778265416622162, "learning_rate": 7.816021270667544e-05, "loss": 0.0233, "step": 34410 }, { "grad_norm": 0.27538469433784485, "learning_rate": 7.81465482531095e-05, "loss": 0.0199, "step": 34420 }, { "grad_norm": 0.15610980987548828, "learning_rate": 7.813288072147938e-05, "loss": 0.0224, "step": 34430 }, { "grad_norm": 0.21462677419185638, "learning_rate": 7.811921011327972e-05, "loss": 0.0243, "step": 34440 }, { "grad_norm": 0.20444513857364655, "learning_rate": 7.810553643000549e-05, "loss": 0.026, "step": 34450 }, { "grad_norm": 0.26526734232902527, "learning_rate": 7.809185967315206e-05, "loss": 0.0226, "step": 34460 }, { "grad_norm": 0.17262716591358185, "learning_rate": 7.80781798442151e-05, "loss": 0.0275, "step": 34470 }, { "grad_norm": 0.3219578266143799, "learning_rate": 7.806449694469061e-05, "loss": 0.0154, "step": 34480 }, { "grad_norm": 0.2153097540140152, "learning_rate": 7.805081097607492e-05, "loss": 0.0269, "step": 34490 }, { "grad_norm": 0.30681419372558594, "learning_rate": 7.803712193986474e-05, "loss": 0.0228, "step": 34500 }, { "grad_norm": 0.21733786165714264, "learning_rate": 7.802342983755702e-05, "loss": 0.0185, "step": 34510 }, { "grad_norm": 0.23509560525417328, "learning_rate": 7.800973467064918e-05, "loss": 0.0263, "step": 34520 }, { "grad_norm": 0.19185468554496765, "learning_rate": 7.799603644063884e-05, "loss": 0.0362, "step": 34530 }, { "grad_norm": 0.22745627164840698, "learning_rate": 7.798233514902405e-05, "loss": 0.022, "step": 34540 }, { "grad_norm": 0.1502811312675476, "learning_rate": 7.796863079730318e-05, "loss": 0.014, "step": 34550 }, { "grad_norm": 0.22238010168075562, "learning_rate": 7.795492338697488e-05, "loss": 0.0225, "step": 34560 }, { "grad_norm": 0.11199017614126205, "learning_rate": 7.794121291953819e-05, "loss": 0.0263, "step": 34570 }, { "grad_norm": 0.17439471185207367, "learning_rate": 7.792749939649246e-05, "loss": 0.0276, "step": 34580 }, { "grad_norm": 0.4828721582889557, "learning_rate": 7.79137828193374e-05, "loss": 0.039, "step": 34590 }, { "grad_norm": 0.48486676812171936, "learning_rate": 7.790006318957301e-05, "loss": 0.0258, "step": 34600 }, { "grad_norm": 0.3131629526615143, "learning_rate": 7.788634050869965e-05, "loss": 0.0353, "step": 34610 }, { "grad_norm": 0.24088677763938904, "learning_rate": 7.787261477821803e-05, "loss": 0.022, "step": 34620 }, { "grad_norm": 0.3518871068954468, "learning_rate": 7.785888599962916e-05, "loss": 0.0203, "step": 34630 }, { "grad_norm": 0.16917914152145386, "learning_rate": 7.784515417443439e-05, "loss": 0.0263, "step": 34640 }, { "grad_norm": 0.5582308173179626, "learning_rate": 7.783141930413545e-05, "loss": 0.029, "step": 34650 }, { "grad_norm": 0.2678965628147125, "learning_rate": 7.78176813902343e-05, "loss": 0.0172, "step": 34660 }, { "grad_norm": 0.20938190817832947, "learning_rate": 7.780394043423336e-05, "loss": 0.0193, "step": 34670 }, { "grad_norm": 0.22532786428928375, "learning_rate": 7.77901964376353e-05, "loss": 0.0235, "step": 34680 }, { "grad_norm": 0.14980114996433258, "learning_rate": 7.777644940194316e-05, "loss": 0.0315, "step": 34690 }, { "grad_norm": 0.35353630781173706, "learning_rate": 7.776269932866023e-05, "loss": 0.0248, "step": 34700 }, { "grad_norm": 0.15954339504241943, "learning_rate": 7.774894621929026e-05, "loss": 0.0329, "step": 34710 }, { "grad_norm": 0.24683311581611633, "learning_rate": 7.773519007533725e-05, "loss": 0.0356, "step": 34720 }, { "grad_norm": 0.2802749574184418, "learning_rate": 7.772143089830556e-05, "loss": 0.0258, "step": 34730 }, { "grad_norm": 0.33879613876342773, "learning_rate": 7.770766868969985e-05, "loss": 0.0217, "step": 34740 }, { "grad_norm": 0.3645053803920746, "learning_rate": 7.769390345102518e-05, "loss": 0.029, "step": 34750 }, { "grad_norm": 0.23187999427318573, "learning_rate": 7.768013518378683e-05, "loss": 0.024, "step": 34760 }, { "grad_norm": 0.6752889156341553, "learning_rate": 7.766636388949053e-05, "loss": 0.0266, "step": 34770 }, { "grad_norm": 0.18535439670085907, "learning_rate": 7.765258956964229e-05, "loss": 0.0286, "step": 34780 }, { "grad_norm": 0.1346304565668106, "learning_rate": 7.76388122257484e-05, "loss": 0.0215, "step": 34790 }, { "grad_norm": 0.15764370560646057, "learning_rate": 7.762503185931558e-05, "loss": 0.0218, "step": 34800 }, { "grad_norm": 0.17647235095500946, "learning_rate": 7.76112484718508e-05, "loss": 0.0231, "step": 34810 }, { "grad_norm": 0.3688848614692688, "learning_rate": 7.75974620648614e-05, "loss": 0.0208, "step": 34820 }, { "grad_norm": 0.28488993644714355, "learning_rate": 7.758367263985503e-05, "loss": 0.0245, "step": 34830 }, { "grad_norm": 0.18265506625175476, "learning_rate": 7.75698801983397e-05, "loss": 0.0245, "step": 34840 }, { "grad_norm": 0.4151705503463745, "learning_rate": 7.755608474182372e-05, "loss": 0.021, "step": 34850 }, { "grad_norm": 0.19796834886074066, "learning_rate": 7.754228627181574e-05, "loss": 0.0309, "step": 34860 }, { "grad_norm": 0.2812718451023102, "learning_rate": 7.752848478982476e-05, "loss": 0.0258, "step": 34870 }, { "grad_norm": 0.45716026425361633, "learning_rate": 7.751468029736006e-05, "loss": 0.0274, "step": 34880 }, { "grad_norm": 0.29146793484687805, "learning_rate": 7.750087279593129e-05, "loss": 0.03, "step": 34890 }, { "grad_norm": 0.34412094950675964, "learning_rate": 7.748706228704843e-05, "loss": 0.0222, "step": 34900 }, { "grad_norm": 0.32799243927001953, "learning_rate": 7.747324877222176e-05, "loss": 0.0341, "step": 34910 }, { "grad_norm": 0.3414035737514496, "learning_rate": 7.745943225296188e-05, "loss": 0.0207, "step": 34920 }, { "grad_norm": 0.20316651463508606, "learning_rate": 7.744561273077981e-05, "loss": 0.0216, "step": 34930 }, { "grad_norm": 0.24421578645706177, "learning_rate": 7.743179020718678e-05, "loss": 0.0203, "step": 34940 }, { "grad_norm": 0.2332359254360199, "learning_rate": 7.741796468369443e-05, "loss": 0.0216, "step": 34950 }, { "grad_norm": 0.3250187933444977, "learning_rate": 7.740413616181466e-05, "loss": 0.0299, "step": 34960 }, { "grad_norm": 0.34126874804496765, "learning_rate": 7.739030464305978e-05, "loss": 0.027, "step": 34970 }, { "grad_norm": 0.11430435627698898, "learning_rate": 7.737647012894235e-05, "loss": 0.0251, "step": 34980 }, { "grad_norm": 0.27989333868026733, "learning_rate": 7.736263262097532e-05, "loss": 0.022, "step": 34990 }, { "grad_norm": 0.17317911982536316, "learning_rate": 7.734879212067192e-05, "loss": 0.025, "step": 35000 }, { "grad_norm": 0.19090506434440613, "learning_rate": 7.733494862954573e-05, "loss": 0.0247, "step": 35010 }, { "grad_norm": 0.528856098651886, "learning_rate": 7.732110214911066e-05, "loss": 0.0248, "step": 35020 }, { "grad_norm": 0.3474862277507782, "learning_rate": 7.730725268088094e-05, "loss": 0.0273, "step": 35030 }, { "grad_norm": 0.3014448583126068, "learning_rate": 7.729340022637112e-05, "loss": 0.0306, "step": 35040 }, { "grad_norm": 0.2960444986820221, "learning_rate": 7.727954478709607e-05, "loss": 0.0268, "step": 35050 }, { "grad_norm": 0.3265894055366516, "learning_rate": 7.726568636457103e-05, "loss": 0.0283, "step": 35060 }, { "grad_norm": 0.5002930164337158, "learning_rate": 7.725182496031153e-05, "loss": 0.0272, "step": 35070 }, { "grad_norm": 0.31043824553489685, "learning_rate": 7.723796057583341e-05, "loss": 0.0219, "step": 35080 }, { "grad_norm": 0.35166704654693604, "learning_rate": 7.722409321265287e-05, "loss": 0.0273, "step": 35090 }, { "grad_norm": 0.2712399661540985, "learning_rate": 7.721022287228645e-05, "loss": 0.039, "step": 35100 }, { "grad_norm": 0.15103229880332947, "learning_rate": 7.719634955625094e-05, "loss": 0.0287, "step": 35110 }, { "grad_norm": 0.19517457485198975, "learning_rate": 7.718247326606355e-05, "loss": 0.0236, "step": 35120 }, { "grad_norm": 0.49828267097473145, "learning_rate": 7.716859400324175e-05, "loss": 0.0244, "step": 35130 }, { "grad_norm": 0.28369563817977905, "learning_rate": 7.715471176930335e-05, "loss": 0.0403, "step": 35140 }, { "grad_norm": 0.23484420776367188, "learning_rate": 7.714082656576651e-05, "loss": 0.0251, "step": 35150 }, { "grad_norm": 0.3983788788318634, "learning_rate": 7.712693839414968e-05, "loss": 0.0305, "step": 35160 }, { "grad_norm": 0.1598626673221588, "learning_rate": 7.711304725597164e-05, "loss": 0.0231, "step": 35170 }, { "grad_norm": 0.2931065261363983, "learning_rate": 7.709915315275151e-05, "loss": 0.0195, "step": 35180 }, { "grad_norm": 0.14483369886875153, "learning_rate": 7.708525608600876e-05, "loss": 0.0314, "step": 35190 }, { "grad_norm": 0.34672659635543823, "learning_rate": 7.707135605726311e-05, "loss": 0.035, "step": 35200 }, { "grad_norm": 0.17755743861198425, "learning_rate": 7.705745306803466e-05, "loss": 0.0485, "step": 35210 }, { "grad_norm": 0.22763733565807343, "learning_rate": 7.704354711984383e-05, "loss": 0.0175, "step": 35220 }, { "grad_norm": 0.29178667068481445, "learning_rate": 7.702963821421133e-05, "loss": 0.0238, "step": 35230 }, { "grad_norm": 0.19107845425605774, "learning_rate": 7.701572635265826e-05, "loss": 0.0152, "step": 35240 }, { "grad_norm": 0.2266652137041092, "learning_rate": 7.700181153670596e-05, "loss": 0.0203, "step": 35250 }, { "grad_norm": 0.11687598377466202, "learning_rate": 7.698789376787616e-05, "loss": 0.0199, "step": 35260 }, { "grad_norm": 0.23166874051094055, "learning_rate": 7.697397304769085e-05, "loss": 0.0218, "step": 35270 }, { "grad_norm": 0.21226420998573303, "learning_rate": 7.696004937767241e-05, "loss": 0.034, "step": 35280 }, { "grad_norm": 0.4755197763442993, "learning_rate": 7.694612275934352e-05, "loss": 0.0315, "step": 35290 }, { "grad_norm": 0.2007126808166504, "learning_rate": 7.693219319422714e-05, "loss": 0.0275, "step": 35300 }, { "grad_norm": 0.2825894057750702, "learning_rate": 7.69182606838466e-05, "loss": 0.0281, "step": 35310 }, { "grad_norm": 0.18242746591567993, "learning_rate": 7.690432522972558e-05, "loss": 0.0243, "step": 35320 }, { "grad_norm": 0.18972253799438477, "learning_rate": 7.689038683338796e-05, "loss": 0.0202, "step": 35330 }, { "grad_norm": 0.20602847635746002, "learning_rate": 7.687644549635808e-05, "loss": 0.0141, "step": 35340 }, { "grad_norm": 0.28448694944381714, "learning_rate": 7.686250122016053e-05, "loss": 0.0316, "step": 35350 }, { "grad_norm": 0.14927247166633606, "learning_rate": 7.684855400632022e-05, "loss": 0.0184, "step": 35360 }, { "grad_norm": 0.18018674850463867, "learning_rate": 7.683460385636243e-05, "loss": 0.0207, "step": 35370 }, { "grad_norm": 0.21098734438419342, "learning_rate": 7.68206507718127e-05, "loss": 0.0286, "step": 35380 }, { "grad_norm": 0.30444926023483276, "learning_rate": 7.680669475419693e-05, "loss": 0.0262, "step": 35390 }, { "grad_norm": 0.20331789553165436, "learning_rate": 7.679273580504132e-05, "loss": 0.0298, "step": 35400 }, { "grad_norm": 0.2349368929862976, "learning_rate": 7.67787739258724e-05, "loss": 0.0289, "step": 35410 }, { "grad_norm": 0.34723785519599915, "learning_rate": 7.676480911821705e-05, "loss": 0.0307, "step": 35420 }, { "grad_norm": 0.2964937686920166, "learning_rate": 7.675084138360238e-05, "loss": 0.0355, "step": 35430 }, { "grad_norm": 0.2459254413843155, "learning_rate": 7.673687072355592e-05, "loss": 0.031, "step": 35440 }, { "grad_norm": 0.24049608409404755, "learning_rate": 7.67228971396055e-05, "loss": 0.0301, "step": 35450 }, { "grad_norm": 0.3443625569343567, "learning_rate": 7.670892063327922e-05, "loss": 0.0293, "step": 35460 }, { "grad_norm": 0.20057669281959534, "learning_rate": 7.669494120610552e-05, "loss": 0.0177, "step": 35470 }, { "grad_norm": 0.2246130108833313, "learning_rate": 7.66809588596132e-05, "loss": 0.0186, "step": 35480 }, { "grad_norm": 0.34234562516212463, "learning_rate": 7.666697359533132e-05, "loss": 0.0236, "step": 35490 }, { "grad_norm": 0.35101771354675293, "learning_rate": 7.665298541478932e-05, "loss": 0.0294, "step": 35500 }, { "grad_norm": 0.2800098955631256, "learning_rate": 7.66389943195169e-05, "loss": 0.0245, "step": 35510 }, { "grad_norm": 0.26342976093292236, "learning_rate": 7.66250003110441e-05, "loss": 0.0284, "step": 35520 }, { "grad_norm": 0.17292748391628265, "learning_rate": 7.66110033909013e-05, "loss": 0.0224, "step": 35530 }, { "grad_norm": 0.20209471881389618, "learning_rate": 7.659700356061918e-05, "loss": 0.027, "step": 35540 }, { "grad_norm": 0.2965260148048401, "learning_rate": 7.658300082172875e-05, "loss": 0.0184, "step": 35550 }, { "grad_norm": 0.1638416051864624, "learning_rate": 7.65689951757613e-05, "loss": 0.0319, "step": 35560 }, { "grad_norm": 0.2966587543487549, "learning_rate": 7.655498662424849e-05, "loss": 0.0275, "step": 35570 }, { "grad_norm": 0.30784446001052856, "learning_rate": 7.654097516872227e-05, "loss": 0.0247, "step": 35580 }, { "grad_norm": 0.3052161931991577, "learning_rate": 7.65269608107149e-05, "loss": 0.0254, "step": 35590 }, { "grad_norm": 0.26142632961273193, "learning_rate": 7.6512943551759e-05, "loss": 0.0298, "step": 35600 }, { "grad_norm": 0.21226288378238678, "learning_rate": 7.649892339338743e-05, "loss": 0.0213, "step": 35610 }, { "grad_norm": 0.16246573626995087, "learning_rate": 7.648490033713344e-05, "loss": 0.0173, "step": 35620 }, { "grad_norm": 0.1766727864742279, "learning_rate": 7.647087438453058e-05, "loss": 0.0223, "step": 35630 }, { "grad_norm": 0.23233090341091156, "learning_rate": 7.64568455371127e-05, "loss": 0.02, "step": 35640 }, { "grad_norm": 0.36725810170173645, "learning_rate": 7.644281379641396e-05, "loss": 0.0265, "step": 35650 }, { "grad_norm": 0.22125563025474548, "learning_rate": 7.642877916396887e-05, "loss": 0.0259, "step": 35660 }, { "grad_norm": 0.1948731541633606, "learning_rate": 7.641474164131221e-05, "loss": 0.0229, "step": 35670 }, { "grad_norm": 0.24485021829605103, "learning_rate": 7.640070122997913e-05, "loss": 0.0265, "step": 35680 }, { "grad_norm": 0.28650975227355957, "learning_rate": 7.638665793150505e-05, "loss": 0.0179, "step": 35690 }, { "grad_norm": 0.23103034496307373, "learning_rate": 7.637261174742574e-05, "loss": 0.0172, "step": 35700 }, { "grad_norm": 0.16713282465934753, "learning_rate": 7.635856267927725e-05, "loss": 0.0164, "step": 35710 }, { "grad_norm": 0.14057938754558563, "learning_rate": 7.634451072859597e-05, "loss": 0.0161, "step": 35720 }, { "grad_norm": 0.3815006613731384, "learning_rate": 7.633045589691863e-05, "loss": 0.0322, "step": 35730 }, { "grad_norm": 0.16216623783111572, "learning_rate": 7.63163981857822e-05, "loss": 0.021, "step": 35740 }, { "grad_norm": 0.3131791353225708, "learning_rate": 7.630233759672403e-05, "loss": 0.0242, "step": 35750 }, { "grad_norm": 0.20403426885604858, "learning_rate": 7.628827413128177e-05, "loss": 0.0264, "step": 35760 }, { "grad_norm": 0.09306361526250839, "learning_rate": 7.627420779099338e-05, "loss": 0.0281, "step": 35770 }, { "grad_norm": 0.281673401594162, "learning_rate": 7.626013857739711e-05, "loss": 0.04, "step": 35780 }, { "grad_norm": 0.16692452132701874, "learning_rate": 7.624606649203158e-05, "loss": 0.0208, "step": 35790 }, { "grad_norm": 0.390246719121933, "learning_rate": 7.623199153643569e-05, "loss": 0.0317, "step": 35800 }, { "grad_norm": 0.28050103783607483, "learning_rate": 7.621791371214863e-05, "loss": 0.0164, "step": 35810 }, { "grad_norm": 0.26280418038368225, "learning_rate": 7.620383302070995e-05, "loss": 0.0183, "step": 35820 }, { "grad_norm": 0.20970851182937622, "learning_rate": 7.61897494636595e-05, "loss": 0.019, "step": 35830 }, { "grad_norm": 0.19489355385303497, "learning_rate": 7.617566304253739e-05, "loss": 0.023, "step": 35840 }, { "grad_norm": 0.2223191112279892, "learning_rate": 7.616157375888416e-05, "loss": 0.0214, "step": 35850 }, { "grad_norm": 0.29265061020851135, "learning_rate": 7.614748161424053e-05, "loss": 0.018, "step": 35860 }, { "grad_norm": 0.2844817042350769, "learning_rate": 7.613338661014763e-05, "loss": 0.0241, "step": 35870 }, { "grad_norm": 0.39039871096611023, "learning_rate": 7.611928874814686e-05, "loss": 0.031, "step": 35880 }, { "grad_norm": 0.31955084204673767, "learning_rate": 7.610518802977996e-05, "loss": 0.0194, "step": 35890 }, { "grad_norm": 0.2249039262533188, "learning_rate": 7.609108445658893e-05, "loss": 0.0139, "step": 35900 }, { "grad_norm": 0.2550232410430908, "learning_rate": 7.607697803011612e-05, "loss": 0.0322, "step": 35910 }, { "grad_norm": 0.21824973821640015, "learning_rate": 7.606286875190421e-05, "loss": 0.0291, "step": 35920 }, { "grad_norm": 0.4102672040462494, "learning_rate": 7.604875662349617e-05, "loss": 0.0225, "step": 35930 }, { "grad_norm": 0.08769748359918594, "learning_rate": 7.603464164643525e-05, "loss": 0.0326, "step": 35940 }, { "grad_norm": 0.11487916111946106, "learning_rate": 7.602052382226507e-05, "loss": 0.0269, "step": 35950 }, { "grad_norm": 0.3940480947494507, "learning_rate": 7.600640315252954e-05, "loss": 0.022, "step": 35960 }, { "grad_norm": 0.21984156966209412, "learning_rate": 7.599227963877284e-05, "loss": 0.0255, "step": 35970 }, { "grad_norm": 0.2186051607131958, "learning_rate": 7.597815328253954e-05, "loss": 0.0259, "step": 35980 }, { "grad_norm": 0.1608084887266159, "learning_rate": 7.596402408537443e-05, "loss": 0.0173, "step": 35990 }, { "grad_norm": 0.12665824592113495, "learning_rate": 7.59498920488227e-05, "loss": 0.0302, "step": 36000 }, { "grad_norm": 0.15539956092834473, "learning_rate": 7.593575717442979e-05, "loss": 0.0221, "step": 36010 }, { "grad_norm": 0.18363158404827118, "learning_rate": 7.592161946374147e-05, "loss": 0.0232, "step": 36020 }, { "grad_norm": 0.29434770345687866, "learning_rate": 7.590747891830381e-05, "loss": 0.0178, "step": 36030 }, { "grad_norm": 0.2282899022102356, "learning_rate": 7.58933355396632e-05, "loss": 0.0265, "step": 36040 }, { "grad_norm": 0.1262339949607849, "learning_rate": 7.587918932936636e-05, "loss": 0.019, "step": 36050 }, { "grad_norm": 0.17041702568531036, "learning_rate": 7.586504028896028e-05, "loss": 0.0256, "step": 36060 }, { "grad_norm": 0.22158445417881012, "learning_rate": 7.585088841999228e-05, "loss": 0.0183, "step": 36070 }, { "grad_norm": 0.3251715302467346, "learning_rate": 7.583673372400999e-05, "loss": 0.0248, "step": 36080 }, { "grad_norm": 0.22548191249370575, "learning_rate": 7.582257620256134e-05, "loss": 0.0189, "step": 36090 }, { "grad_norm": 0.327272891998291, "learning_rate": 7.580841585719458e-05, "loss": 0.0282, "step": 36100 }, { "grad_norm": 0.13799597322940826, "learning_rate": 7.579425268945825e-05, "loss": 0.0147, "step": 36110 }, { "grad_norm": 0.17405462265014648, "learning_rate": 7.578008670090127e-05, "loss": 0.0137, "step": 36120 }, { "grad_norm": 0.2568184733390808, "learning_rate": 7.576591789307272e-05, "loss": 0.0289, "step": 36130 }, { "grad_norm": 0.29248496890068054, "learning_rate": 7.575174626752216e-05, "loss": 0.0238, "step": 36140 }, { "grad_norm": 0.15029273927211761, "learning_rate": 7.573757182579934e-05, "loss": 0.0147, "step": 36150 }, { "grad_norm": 0.2171955555677414, "learning_rate": 7.572339456945435e-05, "loss": 0.0268, "step": 36160 }, { "grad_norm": 0.43548959493637085, "learning_rate": 7.570921450003762e-05, "loss": 0.0308, "step": 36170 }, { "grad_norm": 0.4597087800502777, "learning_rate": 7.569503161909983e-05, "loss": 0.0301, "step": 36180 }, { "grad_norm": 0.23338356614112854, "learning_rate": 7.568084592819202e-05, "loss": 0.0262, "step": 36190 }, { "grad_norm": 0.3184359669685364, "learning_rate": 7.566665742886551e-05, "loss": 0.0152, "step": 36200 }, { "grad_norm": 0.36795610189437866, "learning_rate": 7.565246612267194e-05, "loss": 0.0223, "step": 36210 }, { "grad_norm": 0.29506373405456543, "learning_rate": 7.563827201116325e-05, "loss": 0.0319, "step": 36220 }, { "grad_norm": 0.1657000184059143, "learning_rate": 7.562407509589164e-05, "loss": 0.027, "step": 36230 }, { "grad_norm": 0.1687191277742386, "learning_rate": 7.560987537840973e-05, "loss": 0.0169, "step": 36240 }, { "grad_norm": 0.17055071890354156, "learning_rate": 7.559567286027036e-05, "loss": 0.0246, "step": 36250 }, { "grad_norm": 0.3059888482093811, "learning_rate": 7.558146754302668e-05, "loss": 0.0243, "step": 36260 }, { "grad_norm": 0.18498781323432922, "learning_rate": 7.556725942823217e-05, "loss": 0.0209, "step": 36270 }, { "grad_norm": 0.29690414667129517, "learning_rate": 7.555304851744061e-05, "loss": 0.0209, "step": 36280 }, { "grad_norm": 0.13643747568130493, "learning_rate": 7.553883481220608e-05, "loss": 0.0367, "step": 36290 }, { "grad_norm": 0.1826513409614563, "learning_rate": 7.552461831408298e-05, "loss": 0.0214, "step": 36300 }, { "grad_norm": 0.2724856436252594, "learning_rate": 7.551039902462599e-05, "loss": 0.0234, "step": 36310 }, { "grad_norm": 0.1580219864845276, "learning_rate": 7.549617694539011e-05, "loss": 0.0195, "step": 36320 }, { "grad_norm": 0.2646006643772125, "learning_rate": 7.548195207793066e-05, "loss": 0.0208, "step": 36330 }, { "grad_norm": 0.8724370002746582, "learning_rate": 7.546772442380323e-05, "loss": 0.0181, "step": 36340 }, { "grad_norm": 0.291513055562973, "learning_rate": 7.545349398456376e-05, "loss": 0.0296, "step": 36350 }, { "grad_norm": 0.18287523090839386, "learning_rate": 7.543926076176845e-05, "loss": 0.032, "step": 36360 }, { "grad_norm": 0.18511506915092468, "learning_rate": 7.542502475697385e-05, "loss": 0.0353, "step": 36370 }, { "grad_norm": 0.2635105848312378, "learning_rate": 7.541078597173675e-05, "loss": 0.0304, "step": 36380 }, { "grad_norm": 0.20184510946273804, "learning_rate": 7.539654440761431e-05, "loss": 0.0262, "step": 36390 }, { "grad_norm": 0.2309504896402359, "learning_rate": 7.538230006616395e-05, "loss": 0.0198, "step": 36400 }, { "grad_norm": 0.23637717962265015, "learning_rate": 7.536805294894342e-05, "loss": 0.0208, "step": 36410 }, { "grad_norm": 0.22097396850585938, "learning_rate": 7.535380305751076e-05, "loss": 0.0238, "step": 36420 }, { "grad_norm": 0.2732923626899719, "learning_rate": 7.533955039342431e-05, "loss": 0.0269, "step": 36430 }, { "grad_norm": 0.17836496233940125, "learning_rate": 7.532529495824274e-05, "loss": 0.0272, "step": 36440 }, { "grad_norm": 0.29300495982170105, "learning_rate": 7.531103675352497e-05, "loss": 0.0313, "step": 36450 }, { "grad_norm": 0.28465116024017334, "learning_rate": 7.529677578083029e-05, "loss": 0.0246, "step": 36460 }, { "grad_norm": 0.14508672058582306, "learning_rate": 7.528251204171823e-05, "loss": 0.0164, "step": 36470 }, { "grad_norm": 0.2526124119758606, "learning_rate": 7.526824553774867e-05, "loss": 0.0262, "step": 36480 }, { "grad_norm": 0.22436518967151642, "learning_rate": 7.525397627048178e-05, "loss": 0.0184, "step": 36490 }, { "grad_norm": 0.1727091521024704, "learning_rate": 7.523970424147802e-05, "loss": 0.0179, "step": 36500 }, { "grad_norm": 0.23497119545936584, "learning_rate": 7.522542945229813e-05, "loss": 0.0235, "step": 36510 }, { "grad_norm": 0.21287725865840912, "learning_rate": 7.521115190450321e-05, "loss": 0.0155, "step": 36520 }, { "grad_norm": 0.19536790251731873, "learning_rate": 7.519687159965462e-05, "loss": 0.025, "step": 36530 }, { "grad_norm": 0.2257952243089676, "learning_rate": 7.518258853931403e-05, "loss": 0.0239, "step": 36540 }, { "grad_norm": 0.28029605746269226, "learning_rate": 7.516830272504342e-05, "loss": 0.0184, "step": 36550 }, { "grad_norm": 0.2852482199668884, "learning_rate": 7.515401415840509e-05, "loss": 0.0332, "step": 36560 }, { "grad_norm": 0.27889737486839294, "learning_rate": 7.513972284096155e-05, "loss": 0.0199, "step": 36570 }, { "grad_norm": 0.34178879857063293, "learning_rate": 7.512542877427574e-05, "loss": 0.0318, "step": 36580 }, { "grad_norm": 0.5281026363372803, "learning_rate": 7.51111319599108e-05, "loss": 0.0273, "step": 36590 }, { "grad_norm": 0.537667453289032, "learning_rate": 7.509683239943024e-05, "loss": 0.0364, "step": 36600 }, { "grad_norm": 0.25572362542152405, "learning_rate": 7.50825300943978e-05, "loss": 0.0268, "step": 36610 }, { "grad_norm": 0.29396653175354004, "learning_rate": 7.50682250463776e-05, "loss": 0.0263, "step": 36620 }, { "grad_norm": 0.20400398969650269, "learning_rate": 7.5053917256934e-05, "loss": 0.0274, "step": 36630 }, { "grad_norm": 0.22151097655296326, "learning_rate": 7.503960672763166e-05, "loss": 0.0243, "step": 36640 }, { "grad_norm": 0.23524266481399536, "learning_rate": 7.502529346003559e-05, "loss": 0.0241, "step": 36650 }, { "grad_norm": 0.2035587877035141, "learning_rate": 7.501097745571106e-05, "loss": 0.0228, "step": 36660 }, { "grad_norm": 0.21318696439266205, "learning_rate": 7.499665871622361e-05, "loss": 0.0325, "step": 36670 }, { "grad_norm": 0.41599443554878235, "learning_rate": 7.498233724313917e-05, "loss": 0.0359, "step": 36680 }, { "grad_norm": 0.22950619459152222, "learning_rate": 7.496801303802387e-05, "loss": 0.0366, "step": 36690 }, { "grad_norm": 0.22657206654548645, "learning_rate": 7.495368610244423e-05, "loss": 0.0181, "step": 36700 }, { "grad_norm": 0.3551487922668457, "learning_rate": 7.493935643796697e-05, "loss": 0.0227, "step": 36710 }, { "grad_norm": 0.38511213660240173, "learning_rate": 7.49250240461592e-05, "loss": 0.0214, "step": 36720 }, { "grad_norm": 0.39274653792381287, "learning_rate": 7.491068892858826e-05, "loss": 0.0204, "step": 36730 }, { "grad_norm": 0.21993385255336761, "learning_rate": 7.489635108682184e-05, "loss": 0.0205, "step": 36740 }, { "grad_norm": 0.3194643259048462, "learning_rate": 7.48820105224279e-05, "loss": 0.0243, "step": 36750 }, { "grad_norm": 0.24092157185077667, "learning_rate": 7.48676672369747e-05, "loss": 0.0189, "step": 36760 }, { "grad_norm": 0.32844701409339905, "learning_rate": 7.485332123203079e-05, "loss": 0.035, "step": 36770 }, { "grad_norm": 0.20626169443130493, "learning_rate": 7.483897250916504e-05, "loss": 0.0351, "step": 36780 }, { "grad_norm": 0.231202632188797, "learning_rate": 7.48246210699466e-05, "loss": 0.0228, "step": 36790 }, { "grad_norm": 0.2068704515695572, "learning_rate": 7.481026691594492e-05, "loss": 0.0201, "step": 36800 }, { "grad_norm": 0.3710997700691223, "learning_rate": 7.479591004872975e-05, "loss": 0.0222, "step": 36810 }, { "grad_norm": 0.25228533148765564, "learning_rate": 7.478155046987114e-05, "loss": 0.031, "step": 36820 }, { "grad_norm": 0.26480135321617126, "learning_rate": 7.476718818093944e-05, "loss": 0.0188, "step": 36830 }, { "grad_norm": 0.3040321469306946, "learning_rate": 7.475282318350527e-05, "loss": 0.0167, "step": 36840 }, { "grad_norm": 0.4139571487903595, "learning_rate": 7.473845547913957e-05, "loss": 0.0216, "step": 36850 }, { "grad_norm": 0.32240042090415955, "learning_rate": 7.472408506941357e-05, "loss": 0.0276, "step": 36860 }, { "grad_norm": 0.12851472198963165, "learning_rate": 7.470971195589881e-05, "loss": 0.0277, "step": 36870 }, { "grad_norm": 0.2376830279827118, "learning_rate": 7.46953361401671e-05, "loss": 0.0258, "step": 36880 }, { "grad_norm": 0.28604555130004883, "learning_rate": 7.468095762379055e-05, "loss": 0.017, "step": 36890 }, { "grad_norm": 0.1860923320055008, "learning_rate": 7.466657640834158e-05, "loss": 0.0282, "step": 36900 }, { "grad_norm": 0.2264956682920456, "learning_rate": 7.46521924953929e-05, "loss": 0.0207, "step": 36910 }, { "grad_norm": 0.25289344787597656, "learning_rate": 7.463780588651752e-05, "loss": 0.0253, "step": 36920 }, { "grad_norm": 0.37895384430885315, "learning_rate": 7.462341658328873e-05, "loss": 0.029, "step": 36930 }, { "grad_norm": 0.13784831762313843, "learning_rate": 7.460902458728012e-05, "loss": 0.0237, "step": 36940 }, { "grad_norm": 0.20429947972297668, "learning_rate": 7.459462990006558e-05, "loss": 0.0152, "step": 36950 }, { "grad_norm": 0.15267877280712128, "learning_rate": 7.458023252321929e-05, "loss": 0.0215, "step": 36960 }, { "grad_norm": 0.4065730571746826, "learning_rate": 7.456583245831573e-05, "loss": 0.0256, "step": 36970 }, { "grad_norm": 0.47316282987594604, "learning_rate": 7.455142970692968e-05, "loss": 0.0309, "step": 36980 }, { "grad_norm": 0.3258224427700043, "learning_rate": 7.453702427063615e-05, "loss": 0.0191, "step": 36990 }, { "grad_norm": 0.24855206906795502, "learning_rate": 7.452261615101057e-05, "loss": 0.0216, "step": 37000 }, { "grad_norm": 0.265656441450119, "learning_rate": 7.450820534962856e-05, "loss": 0.0288, "step": 37010 }, { "grad_norm": 0.19228453934192657, "learning_rate": 7.449379186806604e-05, "loss": 0.0211, "step": 37020 }, { "grad_norm": 0.16868682205677032, "learning_rate": 7.44793757078993e-05, "loss": 0.0226, "step": 37030 }, { "grad_norm": 0.4089629054069519, "learning_rate": 7.44649568707048e-05, "loss": 0.0248, "step": 37040 }, { "grad_norm": 0.21370631456375122, "learning_rate": 7.445053535805942e-05, "loss": 0.0262, "step": 37050 }, { "grad_norm": 0.18316185474395752, "learning_rate": 7.443611117154024e-05, "loss": 0.0214, "step": 37060 }, { "grad_norm": 0.39576491713523865, "learning_rate": 7.442168431272469e-05, "loss": 0.0233, "step": 37070 }, { "grad_norm": 0.18592937290668488, "learning_rate": 7.440725478319044e-05, "loss": 0.0245, "step": 37080 }, { "grad_norm": 0.28933557868003845, "learning_rate": 7.439282258451553e-05, "loss": 0.0181, "step": 37090 }, { "grad_norm": 0.25690290331840515, "learning_rate": 7.43783877182782e-05, "loss": 0.0253, "step": 37100 }, { "grad_norm": 0.1464506983757019, "learning_rate": 7.436395018605707e-05, "loss": 0.0191, "step": 37110 }, { "grad_norm": 0.2569080889225006, "learning_rate": 7.434950998943094e-05, "loss": 0.0213, "step": 37120 }, { "grad_norm": 0.23228970170021057, "learning_rate": 7.433506712997902e-05, "loss": 0.0273, "step": 37130 }, { "grad_norm": 0.21123604476451874, "learning_rate": 7.432062160928077e-05, "loss": 0.029, "step": 37140 }, { "grad_norm": 0.25284141302108765, "learning_rate": 7.430617342891588e-05, "loss": 0.0167, "step": 37150 }, { "grad_norm": 0.19029183685779572, "learning_rate": 7.429172259046442e-05, "loss": 0.023, "step": 37160 }, { "grad_norm": 0.2439582645893097, "learning_rate": 7.42772690955067e-05, "loss": 0.0245, "step": 37170 }, { "grad_norm": 0.226533442735672, "learning_rate": 7.426281294562332e-05, "loss": 0.0408, "step": 37180 }, { "grad_norm": 0.1635388284921646, "learning_rate": 7.424835414239519e-05, "loss": 0.0247, "step": 37190 }, { "grad_norm": 1.2807602882385254, "learning_rate": 7.423389268740352e-05, "loss": 0.022, "step": 37200 }, { "grad_norm": 0.22140783071517944, "learning_rate": 7.42194285822298e-05, "loss": 0.0232, "step": 37210 }, { "grad_norm": 0.14054571092128754, "learning_rate": 7.420496182845576e-05, "loss": 0.0181, "step": 37220 }, { "grad_norm": 0.3306823670864105, "learning_rate": 7.419049242766348e-05, "loss": 0.0364, "step": 37230 }, { "grad_norm": 0.24792078137397766, "learning_rate": 7.417602038143533e-05, "loss": 0.0252, "step": 37240 }, { "grad_norm": 0.5902549028396606, "learning_rate": 7.416154569135393e-05, "loss": 0.0315, "step": 37250 }, { "grad_norm": 0.44164615869522095, "learning_rate": 7.414706835900222e-05, "loss": 0.0146, "step": 37260 }, { "grad_norm": 0.3615318238735199, "learning_rate": 7.413258838596345e-05, "loss": 0.0204, "step": 37270 }, { "grad_norm": 0.11204084753990173, "learning_rate": 7.411810577382106e-05, "loss": 0.0255, "step": 37280 }, { "grad_norm": 0.26166433095932007, "learning_rate": 7.410362052415889e-05, "loss": 0.0281, "step": 37290 }, { "grad_norm": 0.4767489731311798, "learning_rate": 7.408913263856102e-05, "loss": 0.024, "step": 37300 }, { "grad_norm": 0.18451261520385742, "learning_rate": 7.407464211861183e-05, "loss": 0.0205, "step": 37310 }, { "grad_norm": 0.16159620881080627, "learning_rate": 7.406014896589597e-05, "loss": 0.0342, "step": 37320 }, { "grad_norm": 0.12777777016162872, "learning_rate": 7.404565318199838e-05, "loss": 0.0251, "step": 37330 }, { "grad_norm": 0.2581344544887543, "learning_rate": 7.403115476850432e-05, "loss": 0.0239, "step": 37340 }, { "grad_norm": 0.26700669527053833, "learning_rate": 7.401665372699932e-05, "loss": 0.0243, "step": 37350 }, { "grad_norm": 0.4817737936973572, "learning_rate": 7.400215005906916e-05, "loss": 0.0281, "step": 37360 }, { "grad_norm": 0.3170481026172638, "learning_rate": 7.398764376629996e-05, "loss": 0.0321, "step": 37370 }, { "grad_norm": 0.349315345287323, "learning_rate": 7.39731348502781e-05, "loss": 0.0242, "step": 37380 }, { "grad_norm": 0.4198109209537506, "learning_rate": 7.395862331259028e-05, "loss": 0.0215, "step": 37390 }, { "grad_norm": 0.3201405704021454, "learning_rate": 7.394410915482341e-05, "loss": 0.0302, "step": 37400 }, { "grad_norm": 0.3575313687324524, "learning_rate": 7.392959237856476e-05, "loss": 0.029, "step": 37410 }, { "grad_norm": 0.22156964242458344, "learning_rate": 7.391507298540187e-05, "loss": 0.024, "step": 37420 }, { "grad_norm": 0.26986104249954224, "learning_rate": 7.390055097692257e-05, "loss": 0.0178, "step": 37430 }, { "grad_norm": 0.3108830749988556, "learning_rate": 7.388602635471494e-05, "loss": 0.0371, "step": 37440 }, { "grad_norm": 0.17244596779346466, "learning_rate": 7.387149912036738e-05, "loss": 0.0393, "step": 37450 }, { "grad_norm": 0.1887505203485489, "learning_rate": 7.385696927546854e-05, "loss": 0.0234, "step": 37460 }, { "grad_norm": 0.2072995901107788, "learning_rate": 7.384243682160743e-05, "loss": 0.032, "step": 37470 }, { "grad_norm": 0.2504875659942627, "learning_rate": 7.382790176037329e-05, "loss": 0.0209, "step": 37480 }, { "grad_norm": 0.3605610728263855, "learning_rate": 7.381336409335562e-05, "loss": 0.022, "step": 37490 }, { "grad_norm": 0.26044392585754395, "learning_rate": 7.379882382214426e-05, "loss": 0.0294, "step": 37500 }, { "grad_norm": 0.25707343220710754, "learning_rate": 7.378428094832931e-05, "loss": 0.0294, "step": 37510 }, { "grad_norm": 0.3237604796886444, "learning_rate": 7.376973547350114e-05, "loss": 0.0316, "step": 37520 }, { "grad_norm": 0.24432934820652008, "learning_rate": 7.375518739925046e-05, "loss": 0.023, "step": 37530 }, { "grad_norm": 0.3714655339717865, "learning_rate": 7.374063672716818e-05, "loss": 0.0305, "step": 37540 }, { "grad_norm": 0.14651478826999664, "learning_rate": 7.372608345884558e-05, "loss": 0.0233, "step": 37550 }, { "grad_norm": 0.5506337285041809, "learning_rate": 7.371152759587415e-05, "loss": 0.0284, "step": 37560 }, { "grad_norm": 0.15750552713871002, "learning_rate": 7.369696913984572e-05, "loss": 0.0172, "step": 37570 }, { "grad_norm": 0.1639295220375061, "learning_rate": 7.368240809235237e-05, "loss": 0.0223, "step": 37580 }, { "grad_norm": 0.26644912362098694, "learning_rate": 7.366784445498649e-05, "loss": 0.0264, "step": 37590 }, { "grad_norm": 0.35723358392715454, "learning_rate": 7.36532782293407e-05, "loss": 0.0265, "step": 37600 }, { "grad_norm": 0.4812975525856018, "learning_rate": 7.363870941700797e-05, "loss": 0.0263, "step": 37610 }, { "grad_norm": 0.118041031062603, "learning_rate": 7.362413801958152e-05, "loss": 0.0232, "step": 37620 }, { "grad_norm": 0.16504772007465363, "learning_rate": 7.360956403865486e-05, "loss": 0.0341, "step": 37630 }, { "grad_norm": 0.3892430365085602, "learning_rate": 7.359498747582177e-05, "loss": 0.0255, "step": 37640 }, { "grad_norm": 0.15047602355480194, "learning_rate": 7.358040833267634e-05, "loss": 0.0205, "step": 37650 }, { "grad_norm": 0.42261698842048645, "learning_rate": 7.356582661081288e-05, "loss": 0.0301, "step": 37660 }, { "grad_norm": 0.5517581105232239, "learning_rate": 7.355124231182607e-05, "loss": 0.0221, "step": 37670 }, { "grad_norm": 0.47468695044517517, "learning_rate": 7.353665543731083e-05, "loss": 0.0227, "step": 37680 }, { "grad_norm": 0.07924138754606247, "learning_rate": 7.352206598886229e-05, "loss": 0.0157, "step": 37690 }, { "grad_norm": 0.17666108906269073, "learning_rate": 7.350747396807601e-05, "loss": 0.0189, "step": 37700 }, { "grad_norm": 0.22204358875751495, "learning_rate": 7.349287937654772e-05, "loss": 0.0136, "step": 37710 }, { "grad_norm": 0.2493535876274109, "learning_rate": 7.347828221587345e-05, "loss": 0.0237, "step": 37720 }, { "grad_norm": 0.300708144903183, "learning_rate": 7.346368248764955e-05, "loss": 0.015, "step": 37730 }, { "grad_norm": 0.5614163279533386, "learning_rate": 7.34490801934726e-05, "loss": 0.0246, "step": 37740 }, { "grad_norm": 0.4184802174568176, "learning_rate": 7.343447533493947e-05, "loss": 0.028, "step": 37750 }, { "grad_norm": 0.31077197194099426, "learning_rate": 7.341986791364739e-05, "loss": 0.037, "step": 37760 }, { "grad_norm": 0.2704346776008606, "learning_rate": 7.340525793119374e-05, "loss": 0.0257, "step": 37770 }, { "grad_norm": 0.3855232000350952, "learning_rate": 7.33906453891763e-05, "loss": 0.0313, "step": 37780 }, { "grad_norm": 0.3830908238887787, "learning_rate": 7.337603028919303e-05, "loss": 0.0249, "step": 37790 }, { "grad_norm": 0.28463560342788696, "learning_rate": 7.336141263284225e-05, "loss": 0.0224, "step": 37800 }, { "grad_norm": 0.3115065097808838, "learning_rate": 7.334679242172249e-05, "loss": 0.0171, "step": 37810 }, { "grad_norm": 0.370399534702301, "learning_rate": 7.33321696574326e-05, "loss": 0.0187, "step": 37820 }, { "grad_norm": 0.21373963356018066, "learning_rate": 7.331754434157174e-05, "loss": 0.0204, "step": 37830 }, { "grad_norm": 0.23182813823223114, "learning_rate": 7.330291647573928e-05, "loss": 0.034, "step": 37840 }, { "grad_norm": 0.20283672213554382, "learning_rate": 7.32882860615349e-05, "loss": 0.0228, "step": 37850 }, { "grad_norm": 0.29733261466026306, "learning_rate": 7.327365310055858e-05, "loss": 0.0243, "step": 37860 }, { "grad_norm": 0.22177189588546753, "learning_rate": 7.325901759441055e-05, "loss": 0.0184, "step": 37870 }, { "grad_norm": 0.2158878892660141, "learning_rate": 7.324437954469132e-05, "loss": 0.0217, "step": 37880 }, { "grad_norm": 0.267912358045578, "learning_rate": 7.32297389530017e-05, "loss": 0.0175, "step": 37890 }, { "grad_norm": 0.26211732625961304, "learning_rate": 7.321509582094274e-05, "loss": 0.0137, "step": 37900 }, { "grad_norm": 0.11083127558231354, "learning_rate": 7.320045015011582e-05, "loss": 0.0175, "step": 37910 }, { "grad_norm": 0.21344833076000214, "learning_rate": 7.318580194212253e-05, "loss": 0.0213, "step": 37920 }, { "grad_norm": 0.21704940497875214, "learning_rate": 7.317115119856483e-05, "loss": 0.0208, "step": 37930 }, { "grad_norm": 0.31795912981033325, "learning_rate": 7.315649792104485e-05, "loss": 0.0317, "step": 37940 }, { "grad_norm": 0.36003103852272034, "learning_rate": 7.314184211116507e-05, "loss": 0.039, "step": 37950 }, { "grad_norm": 0.2693921625614166, "learning_rate": 7.312718377052823e-05, "loss": 0.029, "step": 37960 }, { "grad_norm": 0.19336429238319397, "learning_rate": 7.311252290073737e-05, "loss": 0.0139, "step": 37970 }, { "grad_norm": 0.188725084066391, "learning_rate": 7.309785950339572e-05, "loss": 0.0174, "step": 37980 }, { "grad_norm": 0.28497186303138733, "learning_rate": 7.308319358010692e-05, "loss": 0.0199, "step": 37990 }, { "grad_norm": 0.11656353622674942, "learning_rate": 7.306852513247474e-05, "loss": 0.0412, "step": 38000 }, { "grad_norm": 0.21681345999240875, "learning_rate": 7.305385416210336e-05, "loss": 0.0299, "step": 38010 }, { "grad_norm": 0.25393736362457275, "learning_rate": 7.303918067059711e-05, "loss": 0.0166, "step": 38020 }, { "grad_norm": 0.1463487595319748, "learning_rate": 7.302450465956075e-05, "loss": 0.0218, "step": 38030 }, { "grad_norm": 0.28760600090026855, "learning_rate": 7.300982613059914e-05, "loss": 0.0242, "step": 38040 }, { "grad_norm": 0.18818514049053192, "learning_rate": 7.299514508531757e-05, "loss": 0.0322, "step": 38050 }, { "grad_norm": 0.2211451530456543, "learning_rate": 7.298046152532148e-05, "loss": 0.025, "step": 38060 }, { "grad_norm": 0.20134666562080383, "learning_rate": 7.296577545221668e-05, "loss": 0.019, "step": 38070 }, { "grad_norm": 0.2881568968296051, "learning_rate": 7.29510868676092e-05, "loss": 0.0309, "step": 38080 }, { "grad_norm": 0.31086015701293945, "learning_rate": 7.293639577310538e-05, "loss": 0.0156, "step": 38090 }, { "grad_norm": 0.2461087852716446, "learning_rate": 7.292170217031179e-05, "loss": 0.0216, "step": 38100 }, { "grad_norm": 0.3426939845085144, "learning_rate": 7.290700606083532e-05, "loss": 0.0241, "step": 38110 }, { "grad_norm": 0.16906976699829102, "learning_rate": 7.28923074462831e-05, "loss": 0.0199, "step": 38120 }, { "grad_norm": 0.37186768651008606, "learning_rate": 7.287760632826255e-05, "loss": 0.0317, "step": 38130 }, { "grad_norm": 0.697823166847229, "learning_rate": 7.286290270838139e-05, "loss": 0.0374, "step": 38140 }, { "grad_norm": 0.24247942864894867, "learning_rate": 7.284819658824756e-05, "loss": 0.0289, "step": 38150 }, { "grad_norm": 0.3714185655117035, "learning_rate": 7.283348796946929e-05, "loss": 0.0291, "step": 38160 }, { "grad_norm": 0.22629964351654053, "learning_rate": 7.28187768536551e-05, "loss": 0.0245, "step": 38170 }, { "grad_norm": 0.147453173995018, "learning_rate": 7.280406324241379e-05, "loss": 0.0227, "step": 38180 }, { "grad_norm": 0.19341979920864105, "learning_rate": 7.278934713735441e-05, "loss": 0.0305, "step": 38190 }, { "grad_norm": 0.19924090802669525, "learning_rate": 7.277462854008629e-05, "loss": 0.03, "step": 38200 }, { "grad_norm": 0.32978975772857666, "learning_rate": 7.275990745221904e-05, "loss": 0.0251, "step": 38210 }, { "grad_norm": 0.1836126148700714, "learning_rate": 7.274518387536252e-05, "loss": 0.024, "step": 38220 }, { "grad_norm": 0.4823421835899353, "learning_rate": 7.273045781112689e-05, "loss": 0.0319, "step": 38230 }, { "grad_norm": 0.186028853058815, "learning_rate": 7.271572926112257e-05, "loss": 0.0246, "step": 38240 }, { "grad_norm": 0.21210189163684845, "learning_rate": 7.270099822696024e-05, "loss": 0.0258, "step": 38250 }, { "grad_norm": 0.2418517917394638, "learning_rate": 7.26862647102509e-05, "loss": 0.0284, "step": 38260 }, { "grad_norm": 0.33024847507476807, "learning_rate": 7.267152871260573e-05, "loss": 0.0345, "step": 38270 }, { "grad_norm": 0.3956882953643799, "learning_rate": 7.265679023563629e-05, "loss": 0.0296, "step": 38280 }, { "grad_norm": 0.15013118088245392, "learning_rate": 7.264204928095433e-05, "loss": 0.0215, "step": 38290 }, { "grad_norm": 0.9007606506347656, "learning_rate": 7.262730585017188e-05, "loss": 0.0275, "step": 38300 }, { "grad_norm": 0.2454080730676651, "learning_rate": 7.261255994490131e-05, "loss": 0.0278, "step": 38310 }, { "grad_norm": 0.1261332929134369, "learning_rate": 7.259781156675517e-05, "loss": 0.0241, "step": 38320 }, { "grad_norm": 0.48997262120246887, "learning_rate": 7.258306071734633e-05, "loss": 0.0256, "step": 38330 }, { "grad_norm": 0.18459917604923248, "learning_rate": 7.256830739828792e-05, "loss": 0.0228, "step": 38340 }, { "grad_norm": 0.09967399388551712, "learning_rate": 7.255355161119336e-05, "loss": 0.0163, "step": 38350 }, { "grad_norm": 0.1143704280257225, "learning_rate": 7.253879335767628e-05, "loss": 0.0175, "step": 38360 }, { "grad_norm": 0.1118301972746849, "learning_rate": 7.252403263935067e-05, "loss": 0.025, "step": 38370 }, { "grad_norm": 0.28934362530708313, "learning_rate": 7.25092694578307e-05, "loss": 0.0263, "step": 38380 }, { "grad_norm": 0.2840026021003723, "learning_rate": 7.249450381473087e-05, "loss": 0.0265, "step": 38390 }, { "grad_norm": 0.21783363819122314, "learning_rate": 7.247973571166593e-05, "loss": 0.0228, "step": 38400 }, { "grad_norm": 0.41657784581184387, "learning_rate": 7.246496515025089e-05, "loss": 0.0205, "step": 38410 }, { "grad_norm": 0.40300145745277405, "learning_rate": 7.245019213210101e-05, "loss": 0.0229, "step": 38420 }, { "grad_norm": 0.24575279653072357, "learning_rate": 7.24354166588319e-05, "loss": 0.0301, "step": 38430 }, { "grad_norm": 0.3364158868789673, "learning_rate": 7.242063873205935e-05, "loss": 0.0286, "step": 38440 }, { "grad_norm": 0.33288294076919556, "learning_rate": 7.240585835339946e-05, "loss": 0.0216, "step": 38450 }, { "grad_norm": 0.1285024732351303, "learning_rate": 7.239107552446857e-05, "loss": 0.0187, "step": 38460 }, { "grad_norm": 0.4843687415122986, "learning_rate": 7.237629024688334e-05, "loss": 0.0275, "step": 38470 }, { "grad_norm": 0.4179088771343231, "learning_rate": 7.236150252226064e-05, "loss": 0.0318, "step": 38480 }, { "grad_norm": 0.1904737502336502, "learning_rate": 7.234671235221765e-05, "loss": 0.0331, "step": 38490 }, { "grad_norm": 0.15042400360107422, "learning_rate": 7.233191973837179e-05, "loss": 0.0243, "step": 38500 }, { "grad_norm": 0.16169187426567078, "learning_rate": 7.231712468234078e-05, "loss": 0.0152, "step": 38510 }, { "grad_norm": 0.13697469234466553, "learning_rate": 7.230232718574254e-05, "loss": 0.0279, "step": 38520 }, { "grad_norm": 0.33022284507751465, "learning_rate": 7.228752725019535e-05, "loss": 0.0181, "step": 38530 }, { "grad_norm": 0.41887450218200684, "learning_rate": 7.227272487731769e-05, "loss": 0.0234, "step": 38540 }, { "grad_norm": 0.29928505420684814, "learning_rate": 7.225792006872831e-05, "loss": 0.0243, "step": 38550 }, { "grad_norm": 0.33774128556251526, "learning_rate": 7.224311282604628e-05, "loss": 0.0302, "step": 38560 }, { "grad_norm": 0.29137492179870605, "learning_rate": 7.222830315089085e-05, "loss": 0.0277, "step": 38570 }, { "grad_norm": 0.17927619814872742, "learning_rate": 7.22134910448816e-05, "loss": 0.0233, "step": 38580 }, { "grad_norm": 0.1883411854505539, "learning_rate": 7.219867650963839e-05, "loss": 0.0244, "step": 38590 }, { "grad_norm": 0.33323928713798523, "learning_rate": 7.21838595467813e-05, "loss": 0.0263, "step": 38600 }, { "grad_norm": 0.1267232745885849, "learning_rate": 7.216904015793067e-05, "loss": 0.0232, "step": 38610 }, { "grad_norm": 0.37022584676742554, "learning_rate": 7.215421834470713e-05, "loss": 0.0207, "step": 38620 }, { "grad_norm": 0.24462804198265076, "learning_rate": 7.213939410873158e-05, "loss": 0.0224, "step": 38630 }, { "grad_norm": 0.5225802659988403, "learning_rate": 7.212456745162519e-05, "loss": 0.0188, "step": 38640 }, { "grad_norm": 0.18087618052959442, "learning_rate": 7.210973837500937e-05, "loss": 0.0161, "step": 38650 }, { "grad_norm": 0.4096614718437195, "learning_rate": 7.209490688050578e-05, "loss": 0.0209, "step": 38660 }, { "grad_norm": 0.29108312726020813, "learning_rate": 7.208007296973641e-05, "loss": 0.0221, "step": 38670 }, { "grad_norm": 0.19942398369312286, "learning_rate": 7.206523664432345e-05, "loss": 0.032, "step": 38680 }, { "grad_norm": 0.1836674064397812, "learning_rate": 7.205039790588939e-05, "loss": 0.0183, "step": 38690 }, { "grad_norm": 0.4570169746875763, "learning_rate": 7.203555675605697e-05, "loss": 0.0294, "step": 38700 }, { "grad_norm": 0.1324991136789322, "learning_rate": 7.202071319644917e-05, "loss": 0.0231, "step": 38710 }, { "grad_norm": 0.25478002429008484, "learning_rate": 7.200586722868932e-05, "loss": 0.0296, "step": 38720 }, { "grad_norm": 0.29098647832870483, "learning_rate": 7.199101885440088e-05, "loss": 0.0279, "step": 38730 }, { "grad_norm": 0.27125006914138794, "learning_rate": 7.197616807520771e-05, "loss": 0.019, "step": 38740 }, { "grad_norm": 0.18162909150123596, "learning_rate": 7.196131489273381e-05, "loss": 0.0178, "step": 38750 }, { "grad_norm": 0.22220253944396973, "learning_rate": 7.194645930860355e-05, "loss": 0.0224, "step": 38760 }, { "grad_norm": 0.17090608179569244, "learning_rate": 7.19316013244415e-05, "loss": 0.0199, "step": 38770 }, { "grad_norm": 0.14395378530025482, "learning_rate": 7.191674094187248e-05, "loss": 0.0226, "step": 38780 }, { "grad_norm": 0.2429329752922058, "learning_rate": 7.190187816252165e-05, "loss": 0.0128, "step": 38790 }, { "grad_norm": 0.15940560400485992, "learning_rate": 7.188701298801435e-05, "loss": 0.0179, "step": 38800 }, { "grad_norm": 0.22604498267173767, "learning_rate": 7.18721454199762e-05, "loss": 0.0278, "step": 38810 }, { "grad_norm": 0.22569186985492706, "learning_rate": 7.185727546003312e-05, "loss": 0.0161, "step": 38820 }, { "grad_norm": 0.20727089047431946, "learning_rate": 7.184240310981126e-05, "loss": 0.0226, "step": 38830 }, { "grad_norm": 0.524324357509613, "learning_rate": 7.182752837093704e-05, "loss": 0.0355, "step": 38840 }, { "grad_norm": 0.16159303486347198, "learning_rate": 7.181265124503711e-05, "loss": 0.0311, "step": 38850 }, { "grad_norm": 0.3784952163696289, "learning_rate": 7.179777173373847e-05, "loss": 0.0226, "step": 38860 }, { "grad_norm": 0.33418214321136475, "learning_rate": 7.178288983866826e-05, "loss": 0.0206, "step": 38870 }, { "grad_norm": 0.1886189728975296, "learning_rate": 7.176800556145397e-05, "loss": 0.0295, "step": 38880 }, { "grad_norm": 0.2603863775730133, "learning_rate": 7.175311890372334e-05, "loss": 0.0183, "step": 38890 }, { "grad_norm": 0.44675910472869873, "learning_rate": 7.17382298671043e-05, "loss": 0.0242, "step": 38900 }, { "grad_norm": 0.3433162569999695, "learning_rate": 7.172333845322515e-05, "loss": 0.0278, "step": 38910 }, { "grad_norm": 0.27374008297920227, "learning_rate": 7.170844466371436e-05, "loss": 0.0315, "step": 38920 }, { "grad_norm": 0.19975996017456055, "learning_rate": 7.16935485002007e-05, "loss": 0.0246, "step": 38930 }, { "grad_norm": 0.20295816659927368, "learning_rate": 7.167864996431319e-05, "loss": 0.0298, "step": 38940 }, { "grad_norm": 0.21990135312080383, "learning_rate": 7.166374905768111e-05, "loss": 0.0178, "step": 38950 }, { "grad_norm": 0.11782462149858475, "learning_rate": 7.164884578193401e-05, "loss": 0.0231, "step": 38960 }, { "grad_norm": 0.21936218440532684, "learning_rate": 7.163394013870168e-05, "loss": 0.0166, "step": 38970 }, { "grad_norm": 0.1302085518836975, "learning_rate": 7.161903212961419e-05, "loss": 0.03, "step": 38980 }, { "grad_norm": 0.17435826361179352, "learning_rate": 7.160412175630183e-05, "loss": 0.0193, "step": 38990 }, { "grad_norm": 0.11738220602273941, "learning_rate": 7.158920902039521e-05, "loss": 0.0179, "step": 39000 }, { "grad_norm": 0.21762940287590027, "learning_rate": 7.157429392352514e-05, "loss": 0.0173, "step": 39010 }, { "grad_norm": 0.23162999749183655, "learning_rate": 7.155937646732275e-05, "loss": 0.0201, "step": 39020 }, { "grad_norm": 0.22729966044425964, "learning_rate": 7.154445665341933e-05, "loss": 0.0256, "step": 39030 }, { "grad_norm": 0.24950632452964783, "learning_rate": 7.152953448344654e-05, "loss": 0.0194, "step": 39040 }, { "grad_norm": 0.12398616969585419, "learning_rate": 7.151460995903624e-05, "loss": 0.0205, "step": 39050 }, { "grad_norm": 0.31019896268844604, "learning_rate": 7.149968308182052e-05, "loss": 0.0237, "step": 39060 }, { "grad_norm": 0.1364627480506897, "learning_rate": 7.14847538534318e-05, "loss": 0.0216, "step": 39070 }, { "grad_norm": 0.1810227930545807, "learning_rate": 7.14698222755027e-05, "loss": 0.0203, "step": 39080 }, { "grad_norm": 0.19671212136745453, "learning_rate": 7.14548883496661e-05, "loss": 0.0205, "step": 39090 }, { "grad_norm": 0.22098097205162048, "learning_rate": 7.143995207755517e-05, "loss": 0.0282, "step": 39100 }, { "grad_norm": 0.293688029050827, "learning_rate": 7.142501346080333e-05, "loss": 0.0213, "step": 39110 }, { "grad_norm": 0.23710478842258453, "learning_rate": 7.141007250104421e-05, "loss": 0.0115, "step": 39120 }, { "grad_norm": 0.2935772240161896, "learning_rate": 7.139512919991176e-05, "loss": 0.0192, "step": 39130 }, { "grad_norm": 0.3509962558746338, "learning_rate": 7.138018355904014e-05, "loss": 0.0224, "step": 39140 }, { "grad_norm": 0.35605987906455994, "learning_rate": 7.13652355800638e-05, "loss": 0.0331, "step": 39150 }, { "grad_norm": 0.18782730400562286, "learning_rate": 7.135028526461743e-05, "loss": 0.026, "step": 39160 }, { "grad_norm": 0.22239673137664795, "learning_rate": 7.133533261433594e-05, "loss": 0.0192, "step": 39170 }, { "grad_norm": 0.3040621876716614, "learning_rate": 7.132037763085457e-05, "loss": 0.0234, "step": 39180 }, { "grad_norm": 0.18983061611652374, "learning_rate": 7.130542031580875e-05, "loss": 0.0204, "step": 39190 }, { "grad_norm": 0.3288887143135071, "learning_rate": 7.12904606708342e-05, "loss": 0.0241, "step": 39200 }, { "grad_norm": 0.23562565445899963, "learning_rate": 7.127549869756687e-05, "loss": 0.0239, "step": 39210 }, { "grad_norm": 0.21317826211452484, "learning_rate": 7.126053439764299e-05, "loss": 0.0187, "step": 39220 }, { "grad_norm": 0.3002227544784546, "learning_rate": 7.124556777269904e-05, "loss": 0.0266, "step": 39230 }, { "grad_norm": 0.18432258069515228, "learning_rate": 7.123059882437174e-05, "loss": 0.0282, "step": 39240 }, { "grad_norm": 0.260931134223938, "learning_rate": 7.121562755429807e-05, "loss": 0.0156, "step": 39250 }, { "grad_norm": 0.2582460641860962, "learning_rate": 7.120065396411528e-05, "loss": 0.0224, "step": 39260 }, { "grad_norm": 0.2900961935520172, "learning_rate": 7.118567805546084e-05, "loss": 0.0182, "step": 39270 }, { "grad_norm": 0.28931039571762085, "learning_rate": 7.117069982997248e-05, "loss": 0.0285, "step": 39280 }, { "grad_norm": 0.16627217829227448, "learning_rate": 7.115571928928825e-05, "loss": 0.0364, "step": 39290 }, { "grad_norm": 0.4102002680301666, "learning_rate": 7.114073643504635e-05, "loss": 0.0337, "step": 39300 }, { "grad_norm": 0.20065931975841522, "learning_rate": 7.11257512688853e-05, "loss": 0.0176, "step": 39310 }, { "grad_norm": 0.27860820293426514, "learning_rate": 7.111076379244384e-05, "loss": 0.0211, "step": 39320 }, { "grad_norm": 0.22212441265583038, "learning_rate": 7.109577400736101e-05, "loss": 0.0212, "step": 39330 }, { "grad_norm": 0.287628173828125, "learning_rate": 7.108078191527605e-05, "loss": 0.0219, "step": 39340 }, { "grad_norm": 0.3351532518863678, "learning_rate": 7.106578751782847e-05, "loss": 0.0318, "step": 39350 }, { "grad_norm": 0.1584835946559906, "learning_rate": 7.105079081665803e-05, "loss": 0.0232, "step": 39360 }, { "grad_norm": 0.14662711322307587, "learning_rate": 7.103579181340476e-05, "loss": 0.0243, "step": 39370 }, { "grad_norm": 0.18807312846183777, "learning_rate": 7.102079050970893e-05, "loss": 0.0265, "step": 39380 }, { "grad_norm": 0.25614920258522034, "learning_rate": 7.100578690721105e-05, "loss": 0.0232, "step": 39390 }, { "grad_norm": 0.4285348951816559, "learning_rate": 7.09907810075519e-05, "loss": 0.0334, "step": 39400 }, { "grad_norm": 0.11756635457277298, "learning_rate": 7.097577281237249e-05, "loss": 0.0354, "step": 39410 }, { "grad_norm": 0.15398328006267548, "learning_rate": 7.09607623233141e-05, "loss": 0.0191, "step": 39420 }, { "grad_norm": 0.39587992429733276, "learning_rate": 7.094574954201827e-05, "loss": 0.0334, "step": 39430 }, { "grad_norm": 0.39512336254119873, "learning_rate": 7.093073447012675e-05, "loss": 0.0307, "step": 39440 }, { "grad_norm": 0.228484645485878, "learning_rate": 7.09157171092816e-05, "loss": 0.0199, "step": 39450 }, { "grad_norm": 0.35029587149620056, "learning_rate": 7.090069746112504e-05, "loss": 0.017, "step": 39460 }, { "grad_norm": 0.15460139513015747, "learning_rate": 7.088567552729965e-05, "loss": 0.0228, "step": 39470 }, { "grad_norm": 0.30200332403182983, "learning_rate": 7.087065130944818e-05, "loss": 0.0252, "step": 39480 }, { "grad_norm": 0.16620613634586334, "learning_rate": 7.085562480921366e-05, "loss": 0.0313, "step": 39490 }, { "grad_norm": 0.18143700063228607, "learning_rate": 7.084059602823937e-05, "loss": 0.0209, "step": 39500 }, { "grad_norm": 0.2192346155643463, "learning_rate": 7.082556496816882e-05, "loss": 0.0203, "step": 39510 }, { "grad_norm": 0.5770390629768372, "learning_rate": 7.081053163064582e-05, "loss": 0.0213, "step": 39520 }, { "grad_norm": 0.270395427942276, "learning_rate": 7.079549601731434e-05, "loss": 0.0217, "step": 39530 }, { "grad_norm": 0.36912885308265686, "learning_rate": 7.07804581298187e-05, "loss": 0.0271, "step": 39540 }, { "grad_norm": 0.22343026101589203, "learning_rate": 7.07654179698034e-05, "loss": 0.0187, "step": 39550 }, { "grad_norm": 0.28235140442848206, "learning_rate": 7.075037553891321e-05, "loss": 0.0271, "step": 39560 }, { "grad_norm": 0.18472713232040405, "learning_rate": 7.073533083879315e-05, "loss": 0.0151, "step": 39570 }, { "grad_norm": 0.3117745816707611, "learning_rate": 7.072028387108849e-05, "loss": 0.0293, "step": 39580 }, { "grad_norm": 0.2048783302307129, "learning_rate": 7.070523463744472e-05, "loss": 0.0323, "step": 39590 }, { "grad_norm": 0.45605042576789856, "learning_rate": 7.069018313950763e-05, "loss": 0.0306, "step": 39600 }, { "grad_norm": 0.17787495255470276, "learning_rate": 7.067512937892322e-05, "loss": 0.0146, "step": 39610 }, { "grad_norm": 0.23771503567695618, "learning_rate": 7.066007335733775e-05, "loss": 0.0189, "step": 39620 }, { "grad_norm": 0.23703578114509583, "learning_rate": 7.064501507639772e-05, "loss": 0.0251, "step": 39630 }, { "grad_norm": 0.11853156983852386, "learning_rate": 7.062995453774987e-05, "loss": 0.0304, "step": 39640 }, { "grad_norm": 0.24022682011127472, "learning_rate": 7.061489174304121e-05, "loss": 0.0385, "step": 39650 }, { "grad_norm": 0.141203373670578, "learning_rate": 7.0599826693919e-05, "loss": 0.0161, "step": 39660 }, { "grad_norm": 0.27460959553718567, "learning_rate": 7.05847593920307e-05, "loss": 0.0313, "step": 39670 }, { "grad_norm": 0.3091018497943878, "learning_rate": 7.056968983902406e-05, "loss": 0.026, "step": 39680 }, { "grad_norm": 0.23237714171409607, "learning_rate": 7.055461803654706e-05, "loss": 0.0277, "step": 39690 }, { "grad_norm": 0.350864052772522, "learning_rate": 7.053954398624794e-05, "loss": 0.018, "step": 39700 }, { "grad_norm": 0.24916164577007294, "learning_rate": 7.052446768977518e-05, "loss": 0.0181, "step": 39710 }, { "grad_norm": 0.17156796157360077, "learning_rate": 7.050938914877748e-05, "loss": 0.0171, "step": 39720 }, { "grad_norm": 0.1614697128534317, "learning_rate": 7.049430836490382e-05, "loss": 0.0338, "step": 39730 }, { "grad_norm": 0.3118986189365387, "learning_rate": 7.047922533980342e-05, "loss": 0.0226, "step": 39740 }, { "grad_norm": 0.31592896580696106, "learning_rate": 7.046414007512571e-05, "loss": 0.0222, "step": 39750 }, { "grad_norm": 0.2294723093509674, "learning_rate": 7.044905257252042e-05, "loss": 0.0297, "step": 39760 }, { "grad_norm": 0.15946759283542633, "learning_rate": 7.04339628336375e-05, "loss": 0.0237, "step": 39770 }, { "grad_norm": 0.2501991093158722, "learning_rate": 7.041887086012711e-05, "loss": 0.0212, "step": 39780 }, { "grad_norm": 0.08256127685308456, "learning_rate": 7.040377665363969e-05, "loss": 0.029, "step": 39790 }, { "grad_norm": 0.5304286479949951, "learning_rate": 7.038868021582594e-05, "loss": 0.0363, "step": 39800 }, { "grad_norm": 0.3336864709854126, "learning_rate": 7.037358154833679e-05, "loss": 0.0203, "step": 39810 }, { "grad_norm": 0.5304980278015137, "learning_rate": 7.035848065282339e-05, "loss": 0.036, "step": 39820 }, { "grad_norm": 0.2881603240966797, "learning_rate": 7.034337753093714e-05, "loss": 0.0164, "step": 39830 }, { "grad_norm": 0.3506583869457245, "learning_rate": 7.032827218432972e-05, "loss": 0.0287, "step": 39840 }, { "grad_norm": 0.18448881804943085, "learning_rate": 7.031316461465302e-05, "loss": 0.0184, "step": 39850 }, { "grad_norm": 0.2730218470096588, "learning_rate": 7.029805482355915e-05, "loss": 0.0221, "step": 39860 }, { "grad_norm": 0.18329213559627533, "learning_rate": 7.028294281270055e-05, "loss": 0.0239, "step": 39870 }, { "grad_norm": 0.3703886866569519, "learning_rate": 7.026782858372982e-05, "loss": 0.0293, "step": 39880 }, { "grad_norm": 0.24660620093345642, "learning_rate": 7.025271213829982e-05, "loss": 0.0246, "step": 39890 }, { "grad_norm": 0.23037034273147583, "learning_rate": 7.023759347806366e-05, "loss": 0.0184, "step": 39900 }, { "grad_norm": 0.19601361453533173, "learning_rate": 7.022247260467472e-05, "loss": 0.0229, "step": 39910 }, { "grad_norm": 0.33057740330696106, "learning_rate": 7.020734951978658e-05, "loss": 0.0245, "step": 39920 }, { "grad_norm": 0.27140140533447266, "learning_rate": 7.019222422505307e-05, "loss": 0.0271, "step": 39930 }, { "grad_norm": 0.2623429596424103, "learning_rate": 7.017709672212827e-05, "loss": 0.0276, "step": 39940 }, { "grad_norm": 0.2682918906211853, "learning_rate": 7.016196701266652e-05, "loss": 0.0253, "step": 39950 }, { "grad_norm": 0.1305902898311615, "learning_rate": 7.014683509832238e-05, "loss": 0.0258, "step": 39960 }, { "grad_norm": 0.1763305962085724, "learning_rate": 7.013170098075063e-05, "loss": 0.029, "step": 39970 }, { "grad_norm": 0.27899956703186035, "learning_rate": 7.011656466160632e-05, "loss": 0.0306, "step": 39980 }, { "grad_norm": 0.3086039423942566, "learning_rate": 7.010142614254475e-05, "loss": 0.0224, "step": 39990 }, { "grad_norm": 0.5683091878890991, "learning_rate": 7.008628542522147e-05, "loss": 0.0284, "step": 40000 }, { "grad_norm": 0.31069692969322205, "learning_rate": 7.007114251129218e-05, "loss": 0.0239, "step": 40010 }, { "grad_norm": 0.12645021080970764, "learning_rate": 7.005599740241293e-05, "loss": 0.0203, "step": 40020 }, { "grad_norm": 0.38905155658721924, "learning_rate": 7.004085010023996e-05, "loss": 0.026, "step": 40030 }, { "grad_norm": 0.4267382025718689, "learning_rate": 7.002570060642976e-05, "loss": 0.0204, "step": 40040 }, { "grad_norm": 0.3526453971862793, "learning_rate": 7.001054892263903e-05, "loss": 0.019, "step": 40050 }, { "grad_norm": 0.36552512645721436, "learning_rate": 6.999539505052477e-05, "loss": 0.0221, "step": 40060 }, { "grad_norm": 0.261664479970932, "learning_rate": 6.998023899174418e-05, "loss": 0.0202, "step": 40070 }, { "grad_norm": 0.14353671669960022, "learning_rate": 6.996508074795467e-05, "loss": 0.0156, "step": 40080 }, { "grad_norm": 0.5134303569793701, "learning_rate": 6.994992032081396e-05, "loss": 0.027, "step": 40090 }, { "grad_norm": 0.3345048725605011, "learning_rate": 6.993475771197995e-05, "loss": 0.0217, "step": 40100 }, { "grad_norm": 0.18351684510707855, "learning_rate": 6.991959292311082e-05, "loss": 0.0192, "step": 40110 }, { "grad_norm": 0.2867642343044281, "learning_rate": 6.990442595586495e-05, "loss": 0.0306, "step": 40120 }, { "grad_norm": 0.1564178764820099, "learning_rate": 6.988925681190098e-05, "loss": 0.0189, "step": 40130 }, { "grad_norm": 0.5415828227996826, "learning_rate": 6.987408549287778e-05, "loss": 0.0259, "step": 40140 }, { "grad_norm": 0.3195822536945343, "learning_rate": 6.985891200045449e-05, "loss": 0.0211, "step": 40150 }, { "grad_norm": 0.19789783656597137, "learning_rate": 6.984373633629045e-05, "loss": 0.0249, "step": 40160 }, { "grad_norm": 0.18807685375213623, "learning_rate": 6.98285585020452e-05, "loss": 0.0155, "step": 40170 }, { "grad_norm": 0.4048394560813904, "learning_rate": 6.981337849937864e-05, "loss": 0.0273, "step": 40180 }, { "grad_norm": 0.22584930062294006, "learning_rate": 6.979819632995078e-05, "loss": 0.0242, "step": 40190 }, { "grad_norm": 0.17722004652023315, "learning_rate": 6.978301199542193e-05, "loss": 0.0193, "step": 40200 }, { "grad_norm": 0.2646046578884125, "learning_rate": 6.976782549745263e-05, "loss": 0.0267, "step": 40210 }, { "grad_norm": 0.30338290333747864, "learning_rate": 6.975263683770367e-05, "loss": 0.0289, "step": 40220 }, { "grad_norm": 0.2803438603878021, "learning_rate": 6.973744601783604e-05, "loss": 0.0214, "step": 40230 }, { "grad_norm": 0.23374700546264648, "learning_rate": 6.972225303951097e-05, "loss": 0.0249, "step": 40240 }, { "grad_norm": 0.2577498257160187, "learning_rate": 6.970705790438998e-05, "loss": 0.0144, "step": 40250 }, { "grad_norm": 0.18270714581012726, "learning_rate": 6.969186061413477e-05, "loss": 0.0175, "step": 40260 }, { "grad_norm": 0.2099357694387436, "learning_rate": 6.967666117040727e-05, "loss": 0.03, "step": 40270 }, { "grad_norm": 0.19605332612991333, "learning_rate": 6.966145957486972e-05, "loss": 0.0166, "step": 40280 }, { "grad_norm": 0.24066874384880066, "learning_rate": 6.964625582918449e-05, "loss": 0.0275, "step": 40290 }, { "grad_norm": 0.15363235771656036, "learning_rate": 6.963104993501425e-05, "loss": 0.0218, "step": 40300 }, { "grad_norm": 0.43294164538383484, "learning_rate": 6.961584189402192e-05, "loss": 0.0268, "step": 40310 }, { "grad_norm": 0.13435101509094238, "learning_rate": 6.96006317078706e-05, "loss": 0.0175, "step": 40320 }, { "grad_norm": 0.32168063521385193, "learning_rate": 6.958541937822367e-05, "loss": 0.0212, "step": 40330 }, { "grad_norm": 0.1973530352115631, "learning_rate": 6.957020490674473e-05, "loss": 0.0174, "step": 40340 }, { "grad_norm": 0.23057104647159576, "learning_rate": 6.95549882950976e-05, "loss": 0.0142, "step": 40350 }, { "grad_norm": 0.26501405239105225, "learning_rate": 6.953976954494635e-05, "loss": 0.017, "step": 40360 }, { "grad_norm": 0.15173327922821045, "learning_rate": 6.952454865795528e-05, "loss": 0.0201, "step": 40370 }, { "grad_norm": 0.47251999378204346, "learning_rate": 6.950932563578892e-05, "loss": 0.023, "step": 40380 }, { "grad_norm": 0.5196945071220398, "learning_rate": 6.949410048011206e-05, "loss": 0.0234, "step": 40390 }, { "grad_norm": 0.18001283705234528, "learning_rate": 6.947887319258966e-05, "loss": 0.0191, "step": 40400 }, { "grad_norm": 0.5928637981414795, "learning_rate": 6.946364377488696e-05, "loss": 0.032, "step": 40410 }, { "grad_norm": 0.3521840274333954, "learning_rate": 6.944841222866947e-05, "loss": 0.0285, "step": 40420 }, { "grad_norm": 0.18289363384246826, "learning_rate": 6.943317855560284e-05, "loss": 0.033, "step": 40430 }, { "grad_norm": 0.2278313934803009, "learning_rate": 6.941794275735302e-05, "loss": 0.023, "step": 40440 }, { "grad_norm": 0.2749641537666321, "learning_rate": 6.94027048355862e-05, "loss": 0.029, "step": 40450 }, { "grad_norm": 0.22238922119140625, "learning_rate": 6.938746479196871e-05, "loss": 0.0269, "step": 40460 }, { "grad_norm": 0.2811989486217499, "learning_rate": 6.937222262816724e-05, "loss": 0.0236, "step": 40470 }, { "grad_norm": 0.13913413882255554, "learning_rate": 6.935697834584865e-05, "loss": 0.0248, "step": 40480 }, { "grad_norm": 0.3221683204174042, "learning_rate": 6.934173194667998e-05, "loss": 0.023, "step": 40490 }, { "grad_norm": 0.3104258179664612, "learning_rate": 6.93264834323286e-05, "loss": 0.0246, "step": 40500 }, { "grad_norm": 0.15448665618896484, "learning_rate": 6.931123280446206e-05, "loss": 0.0266, "step": 40510 }, { "grad_norm": 0.30453193187713623, "learning_rate": 6.929598006474811e-05, "loss": 0.0222, "step": 40520 }, { "grad_norm": 0.11042369157075882, "learning_rate": 6.92807252148548e-05, "loss": 0.0149, "step": 40530 }, { "grad_norm": 0.13255801796913147, "learning_rate": 6.926546825645037e-05, "loss": 0.023, "step": 40540 }, { "grad_norm": 0.4513574242591858, "learning_rate": 6.92502091912033e-05, "loss": 0.0178, "step": 40550 }, { "grad_norm": 0.2191835194826126, "learning_rate": 6.923494802078229e-05, "loss": 0.0251, "step": 40560 }, { "grad_norm": 0.34341302514076233, "learning_rate": 6.921968474685628e-05, "loss": 0.0208, "step": 40570 }, { "grad_norm": 0.4814600646495819, "learning_rate": 6.920441937109445e-05, "loss": 0.0216, "step": 40580 }, { "grad_norm": 0.31840524077415466, "learning_rate": 6.91891518951662e-05, "loss": 0.0244, "step": 40590 }, { "grad_norm": 0.307695597410202, "learning_rate": 6.917388232074114e-05, "loss": 0.0341, "step": 40600 }, { "grad_norm": 0.22564692795276642, "learning_rate": 6.915861064948914e-05, "loss": 0.018, "step": 40610 }, { "grad_norm": 0.17909419536590576, "learning_rate": 6.914333688308025e-05, "loss": 0.0263, "step": 40620 }, { "grad_norm": 0.32595908641815186, "learning_rate": 6.912806102318487e-05, "loss": 0.0233, "step": 40630 }, { "grad_norm": 0.2813144624233246, "learning_rate": 6.911278307147347e-05, "loss": 0.0289, "step": 40640 }, { "grad_norm": 0.14904935657978058, "learning_rate": 6.909750302961684e-05, "loss": 0.0201, "step": 40650 }, { "grad_norm": 0.19787608087062836, "learning_rate": 6.9082220899286e-05, "loss": 0.0266, "step": 40660 }, { "grad_norm": 0.12041979283094406, "learning_rate": 6.906693668215219e-05, "loss": 0.0176, "step": 40670 }, { "grad_norm": 0.22039495408535004, "learning_rate": 6.905165037988683e-05, "loss": 0.0181, "step": 40680 }, { "grad_norm": 0.20974716544151306, "learning_rate": 6.903636199416164e-05, "loss": 0.0201, "step": 40690 }, { "grad_norm": 0.13274942338466644, "learning_rate": 6.902107152664851e-05, "loss": 0.0218, "step": 40700 }, { "grad_norm": 0.14891326427459717, "learning_rate": 6.90057789790196e-05, "loss": 0.0195, "step": 40710 }, { "grad_norm": 0.16133743524551392, "learning_rate": 6.899048435294728e-05, "loss": 0.0247, "step": 40720 }, { "grad_norm": 0.24514710903167725, "learning_rate": 6.897518765010415e-05, "loss": 0.0294, "step": 40730 }, { "grad_norm": 0.38509082794189453, "learning_rate": 6.895988887216302e-05, "loss": 0.0244, "step": 40740 }, { "grad_norm": 0.3262447118759155, "learning_rate": 6.894458802079694e-05, "loss": 0.0258, "step": 40750 }, { "grad_norm": 0.1160195916891098, "learning_rate": 6.892928509767922e-05, "loss": 0.0142, "step": 40760 }, { "grad_norm": 0.1590275764465332, "learning_rate": 6.891398010448333e-05, "loss": 0.0219, "step": 40770 }, { "grad_norm": 0.26477837562561035, "learning_rate": 6.8898673042883e-05, "loss": 0.0306, "step": 40780 }, { "grad_norm": 0.34768804907798767, "learning_rate": 6.888336391455222e-05, "loss": 0.0213, "step": 40790 }, { "grad_norm": 0.11916277557611465, "learning_rate": 6.886805272116513e-05, "loss": 0.0182, "step": 40800 }, { "grad_norm": 0.35730454325675964, "learning_rate": 6.885273946439617e-05, "loss": 0.0326, "step": 40810 }, { "grad_norm": 0.3339405059814453, "learning_rate": 6.883742414591998e-05, "loss": 0.0256, "step": 40820 }, { "grad_norm": 0.22084692120552063, "learning_rate": 6.88221067674114e-05, "loss": 0.0199, "step": 40830 }, { "grad_norm": 0.16500157117843628, "learning_rate": 6.88067873305455e-05, "loss": 0.0212, "step": 40840 }, { "grad_norm": 0.13045445084571838, "learning_rate": 6.879146583699765e-05, "loss": 0.0151, "step": 40850 }, { "grad_norm": 0.16318988800048828, "learning_rate": 6.877614228844334e-05, "loss": 0.0231, "step": 40860 }, { "grad_norm": 0.17500197887420654, "learning_rate": 6.876081668655832e-05, "loss": 0.0308, "step": 40870 }, { "grad_norm": 0.5052859783172607, "learning_rate": 6.874548903301861e-05, "loss": 0.0243, "step": 40880 }, { "grad_norm": 0.24633340537548065, "learning_rate": 6.873015932950039e-05, "loss": 0.0198, "step": 40890 }, { "grad_norm": 0.40459710359573364, "learning_rate": 6.871482757768012e-05, "loss": 0.0276, "step": 40900 }, { "grad_norm": 0.23003408312797546, "learning_rate": 6.869949377923442e-05, "loss": 0.0226, "step": 40910 }, { "grad_norm": 0.19843655824661255, "learning_rate": 6.868415793584022e-05, "loss": 0.0178, "step": 40920 }, { "grad_norm": 0.17599281668663025, "learning_rate": 6.86688200491746e-05, "loss": 0.0225, "step": 40930 }, { "grad_norm": 0.21484889090061188, "learning_rate": 6.865348012091488e-05, "loss": 0.0212, "step": 40940 }, { "grad_norm": 0.16818052530288696, "learning_rate": 6.86381381527386e-05, "loss": 0.0234, "step": 40950 }, { "grad_norm": 0.11224786937236786, "learning_rate": 6.862279414632359e-05, "loss": 0.0168, "step": 40960 }, { "grad_norm": 0.4589090943336487, "learning_rate": 6.860744810334778e-05, "loss": 0.0208, "step": 40970 }, { "grad_norm": 0.17257684469223022, "learning_rate": 6.859210002548943e-05, "loss": 0.0166, "step": 40980 }, { "grad_norm": 0.20426811277866364, "learning_rate": 6.8576749914427e-05, "loss": 0.0269, "step": 40990 }, { "grad_norm": 0.2104112207889557, "learning_rate": 6.85613977718391e-05, "loss": 0.0235, "step": 41000 }, { "grad_norm": 0.14440155029296875, "learning_rate": 6.854604359940464e-05, "loss": 0.0169, "step": 41010 }, { "grad_norm": 0.20406007766723633, "learning_rate": 6.853068739880276e-05, "loss": 0.0203, "step": 41020 }, { "grad_norm": 0.36621084809303284, "learning_rate": 6.851532917171276e-05, "loss": 0.0317, "step": 41030 }, { "grad_norm": 0.1840391755104065, "learning_rate": 6.849996891981418e-05, "loss": 0.0276, "step": 41040 }, { "grad_norm": 0.11208371818065643, "learning_rate": 6.848460664478684e-05, "loss": 0.0217, "step": 41050 }, { "grad_norm": 0.28257909417152405, "learning_rate": 6.846924234831069e-05, "loss": 0.0224, "step": 41060 }, { "grad_norm": 0.3132653534412384, "learning_rate": 6.845387603206597e-05, "loss": 0.0172, "step": 41070 }, { "grad_norm": 0.2977137863636017, "learning_rate": 6.843850769773312e-05, "loss": 0.0133, "step": 41080 }, { "grad_norm": 0.24087640643119812, "learning_rate": 6.842313734699278e-05, "loss": 0.0333, "step": 41090 }, { "grad_norm": 0.1827864646911621, "learning_rate": 6.840776498152584e-05, "loss": 0.0165, "step": 41100 }, { "grad_norm": 0.3235110640525818, "learning_rate": 6.83923906030134e-05, "loss": 0.0211, "step": 41110 }, { "grad_norm": 0.4210694134235382, "learning_rate": 6.837701421313677e-05, "loss": 0.0349, "step": 41120 }, { "grad_norm": 0.16986122727394104, "learning_rate": 6.83616358135775e-05, "loss": 0.0208, "step": 41130 }, { "grad_norm": 0.3050192594528198, "learning_rate": 6.834625540601734e-05, "loss": 0.0211, "step": 41140 }, { "grad_norm": 0.1858772337436676, "learning_rate": 6.833087299213829e-05, "loss": 0.021, "step": 41150 }, { "grad_norm": 0.1551387459039688, "learning_rate": 6.83154885736225e-05, "loss": 0.0225, "step": 41160 }, { "grad_norm": 0.26924610137939453, "learning_rate": 6.830010215215244e-05, "loss": 0.0322, "step": 41170 }, { "grad_norm": 0.25232023000717163, "learning_rate": 6.828471372941072e-05, "loss": 0.0198, "step": 41180 }, { "grad_norm": 0.849018931388855, "learning_rate": 6.826932330708018e-05, "loss": 0.0292, "step": 41190 }, { "grad_norm": 0.2005445510149002, "learning_rate": 6.825393088684393e-05, "loss": 0.0212, "step": 41200 }, { "grad_norm": 0.2974821925163269, "learning_rate": 6.823853647038524e-05, "loss": 0.0213, "step": 41210 }, { "grad_norm": 0.350709468126297, "learning_rate": 6.822314005938762e-05, "loss": 0.021, "step": 41220 }, { "grad_norm": 0.2480877786874771, "learning_rate": 6.820774165553479e-05, "loss": 0.0288, "step": 41230 }, { "grad_norm": 0.1642449051141739, "learning_rate": 6.819234126051072e-05, "loss": 0.0233, "step": 41240 }, { "grad_norm": 0.23285095393657684, "learning_rate": 6.817693887599956e-05, "loss": 0.0198, "step": 41250 }, { "grad_norm": 0.27513575553894043, "learning_rate": 6.816153450368568e-05, "loss": 0.0276, "step": 41260 }, { "grad_norm": 0.18501773476600647, "learning_rate": 6.814612814525372e-05, "loss": 0.032, "step": 41270 }, { "grad_norm": 0.2976035177707672, "learning_rate": 6.813071980238846e-05, "loss": 0.018, "step": 41280 }, { "grad_norm": 0.2932951748371124, "learning_rate": 6.811530947677492e-05, "loss": 0.0179, "step": 41290 }, { "grad_norm": 0.4626489281654358, "learning_rate": 6.809989717009839e-05, "loss": 0.0181, "step": 41300 }, { "grad_norm": 0.13895143568515778, "learning_rate": 6.808448288404431e-05, "loss": 0.0284, "step": 41310 }, { "grad_norm": 0.17519322037696838, "learning_rate": 6.806906662029838e-05, "loss": 0.0211, "step": 41320 }, { "grad_norm": 0.15866290032863617, "learning_rate": 6.805364838054647e-05, "loss": 0.0135, "step": 41330 }, { "grad_norm": 0.32107123732566833, "learning_rate": 6.803822816647474e-05, "loss": 0.0188, "step": 41340 }, { "grad_norm": 0.5793605446815491, "learning_rate": 6.802280597976949e-05, "loss": 0.0199, "step": 41350 }, { "grad_norm": 0.33484482765197754, "learning_rate": 6.800738182211727e-05, "loss": 0.0197, "step": 41360 }, { "grad_norm": 0.17068906128406525, "learning_rate": 6.799195569520485e-05, "loss": 0.0307, "step": 41370 }, { "grad_norm": 0.42696499824523926, "learning_rate": 6.79765276007192e-05, "loss": 0.0259, "step": 41380 }, { "grad_norm": 0.32184457778930664, "learning_rate": 6.796109754034753e-05, "loss": 0.0228, "step": 41390 }, { "grad_norm": 0.5140867829322815, "learning_rate": 6.794566551577724e-05, "loss": 0.0318, "step": 41400 }, { "grad_norm": 0.2694591283798218, "learning_rate": 6.793023152869594e-05, "loss": 0.021, "step": 41410 }, { "grad_norm": 0.15104341506958008, "learning_rate": 6.791479558079147e-05, "loss": 0.0152, "step": 41420 }, { "grad_norm": 0.24326784908771515, "learning_rate": 6.789935767375191e-05, "loss": 0.017, "step": 41430 }, { "grad_norm": 0.26591795682907104, "learning_rate": 6.78839178092655e-05, "loss": 0.0134, "step": 41440 }, { "grad_norm": 0.2591296136379242, "learning_rate": 6.786847598902072e-05, "loss": 0.0288, "step": 41450 }, { "grad_norm": 0.14711301028728485, "learning_rate": 6.785303221470627e-05, "loss": 0.0158, "step": 41460 }, { "grad_norm": 0.2544407546520233, "learning_rate": 6.783758648801108e-05, "loss": 0.0185, "step": 41470 }, { "grad_norm": 0.2333642840385437, "learning_rate": 6.782213881062422e-05, "loss": 0.0288, "step": 41480 }, { "grad_norm": 0.32504621148109436, "learning_rate": 6.780668918423508e-05, "loss": 0.0145, "step": 41490 }, { "grad_norm": 0.23189233243465424, "learning_rate": 6.779123761053317e-05, "loss": 0.0218, "step": 41500 }, { "grad_norm": 0.2999033033847809, "learning_rate": 6.777578409120826e-05, "loss": 0.0187, "step": 41510 }, { "grad_norm": 0.2633282244205475, "learning_rate": 6.776032862795033e-05, "loss": 0.0208, "step": 41520 }, { "grad_norm": 0.36785000562667847, "learning_rate": 6.774487122244956e-05, "loss": 0.0298, "step": 41530 }, { "grad_norm": 0.18527188897132874, "learning_rate": 6.772941187639637e-05, "loss": 0.0149, "step": 41540 }, { "grad_norm": 0.3257482945919037, "learning_rate": 6.771395059148134e-05, "loss": 0.015, "step": 41550 }, { "grad_norm": 0.13815389573574066, "learning_rate": 6.76984873693953e-05, "loss": 0.0188, "step": 41560 }, { "grad_norm": 0.41718465089797974, "learning_rate": 6.768302221182931e-05, "loss": 0.0215, "step": 41570 }, { "grad_norm": 0.30690962076187134, "learning_rate": 6.766755512047457e-05, "loss": 0.0228, "step": 41580 }, { "grad_norm": 0.28734466433525085, "learning_rate": 6.765208609702259e-05, "loss": 0.0188, "step": 41590 }, { "grad_norm": 0.1659838706254959, "learning_rate": 6.763661514316499e-05, "loss": 0.0127, "step": 41600 }, { "grad_norm": 0.20467223227024078, "learning_rate": 6.76211422605937e-05, "loss": 0.0311, "step": 41610 }, { "grad_norm": 0.21961086988449097, "learning_rate": 6.760566745100076e-05, "loss": 0.019, "step": 41620 }, { "grad_norm": 0.17691446840763092, "learning_rate": 6.75901907160785e-05, "loss": 0.0181, "step": 41630 }, { "grad_norm": 0.3204704225063324, "learning_rate": 6.757471205751943e-05, "loss": 0.022, "step": 41640 }, { "grad_norm": 0.3126456141471863, "learning_rate": 6.75592314770163e-05, "loss": 0.0188, "step": 41650 }, { "grad_norm": 0.14397498965263367, "learning_rate": 6.7543748976262e-05, "loss": 0.0213, "step": 41660 }, { "grad_norm": 0.1587536334991455, "learning_rate": 6.752826455694968e-05, "loss": 0.0205, "step": 41670 }, { "grad_norm": 0.259595662355423, "learning_rate": 6.751277822077271e-05, "loss": 0.0252, "step": 41680 }, { "grad_norm": 0.3688984811306, "learning_rate": 6.749728996942465e-05, "loss": 0.0178, "step": 41690 }, { "grad_norm": 0.2302970439195633, "learning_rate": 6.748179980459924e-05, "loss": 0.022, "step": 41700 }, { "grad_norm": 0.1890222281217575, "learning_rate": 6.746630772799052e-05, "loss": 0.0237, "step": 41710 }, { "grad_norm": 0.2420598864555359, "learning_rate": 6.745081374129262e-05, "loss": 0.0246, "step": 41720 }, { "grad_norm": 0.2551169693470001, "learning_rate": 6.743531784619998e-05, "loss": 0.0283, "step": 41730 }, { "grad_norm": 0.24190472066402435, "learning_rate": 6.741982004440719e-05, "loss": 0.0224, "step": 41740 }, { "grad_norm": 0.2751319110393524, "learning_rate": 6.740432033760907e-05, "loss": 0.0244, "step": 41750 }, { "grad_norm": 0.1989232450723648, "learning_rate": 6.738881872750066e-05, "loss": 0.0242, "step": 41760 }, { "grad_norm": 0.1084185391664505, "learning_rate": 6.737331521577715e-05, "loss": 0.0193, "step": 41770 }, { "grad_norm": 0.2972572147846222, "learning_rate": 6.735780980413403e-05, "loss": 0.0154, "step": 41780 }, { "grad_norm": 0.29487016797065735, "learning_rate": 6.734230249426692e-05, "loss": 0.0234, "step": 41790 }, { "grad_norm": 0.27973002195358276, "learning_rate": 6.732679328787168e-05, "loss": 0.02, "step": 41800 }, { "grad_norm": 0.07794999331235886, "learning_rate": 6.731128218664438e-05, "loss": 0.0243, "step": 41810 }, { "grad_norm": 0.3738729953765869, "learning_rate": 6.729576919228129e-05, "loss": 0.0173, "step": 41820 }, { "grad_norm": 0.2725837230682373, "learning_rate": 6.728025430647888e-05, "loss": 0.0322, "step": 41830 }, { "grad_norm": 0.20306986570358276, "learning_rate": 6.726473753093383e-05, "loss": 0.0191, "step": 41840 }, { "grad_norm": 0.20359840989112854, "learning_rate": 6.724921886734305e-05, "loss": 0.0239, "step": 41850 }, { "grad_norm": 0.3397493362426758, "learning_rate": 6.723369831740363e-05, "loss": 0.0298, "step": 41860 }, { "grad_norm": 0.19308097660541534, "learning_rate": 6.721817588281288e-05, "loss": 0.0209, "step": 41870 }, { "grad_norm": 0.20664405822753906, "learning_rate": 6.720265156526828e-05, "loss": 0.026, "step": 41880 }, { "grad_norm": 0.22458091378211975, "learning_rate": 6.718712536646758e-05, "loss": 0.0187, "step": 41890 }, { "grad_norm": 0.2877010107040405, "learning_rate": 6.71715972881087e-05, "loss": 0.0295, "step": 41900 }, { "grad_norm": 0.15414369106292725, "learning_rate": 6.715606733188974e-05, "loss": 0.0191, "step": 41910 }, { "grad_norm": 0.21856845915317535, "learning_rate": 6.714053549950909e-05, "loss": 0.016, "step": 41920 }, { "grad_norm": 0.18553249537944794, "learning_rate": 6.71250017926652e-05, "loss": 0.0297, "step": 41930 }, { "grad_norm": 0.3050059974193573, "learning_rate": 6.71094662130569e-05, "loss": 0.0228, "step": 41940 }, { "grad_norm": 0.1732366383075714, "learning_rate": 6.709392876238307e-05, "loss": 0.0163, "step": 41950 }, { "grad_norm": 0.27014660835266113, "learning_rate": 6.707838944234289e-05, "loss": 0.0229, "step": 41960 }, { "grad_norm": 0.13311505317687988, "learning_rate": 6.706284825463574e-05, "loss": 0.0243, "step": 41970 }, { "grad_norm": 0.3029142916202545, "learning_rate": 6.704730520096114e-05, "loss": 0.0227, "step": 41980 }, { "grad_norm": 0.23305408656597137, "learning_rate": 6.703176028301888e-05, "loss": 0.0239, "step": 41990 }, { "grad_norm": 0.46648356318473816, "learning_rate": 6.701621350250892e-05, "loss": 0.021, "step": 42000 }, { "grad_norm": 0.21434757113456726, "learning_rate": 6.700066486113144e-05, "loss": 0.0158, "step": 42010 }, { "grad_norm": 0.22847282886505127, "learning_rate": 6.69851143605868e-05, "loss": 0.0202, "step": 42020 }, { "grad_norm": 0.23392918705940247, "learning_rate": 6.696956200257561e-05, "loss": 0.029, "step": 42030 }, { "grad_norm": 0.1446438431739807, "learning_rate": 6.695400778879863e-05, "loss": 0.0327, "step": 42040 }, { "grad_norm": 0.4660889506340027, "learning_rate": 6.693845172095683e-05, "loss": 0.0342, "step": 42050 }, { "grad_norm": 0.22913405299186707, "learning_rate": 6.692289380075142e-05, "loss": 0.0234, "step": 42060 }, { "grad_norm": 0.16398444771766663, "learning_rate": 6.690733402988379e-05, "loss": 0.0213, "step": 42070 }, { "grad_norm": 0.3404524624347687, "learning_rate": 6.689177241005553e-05, "loss": 0.0218, "step": 42080 }, { "grad_norm": 0.3362029790878296, "learning_rate": 6.687620894296844e-05, "loss": 0.0277, "step": 42090 }, { "grad_norm": 0.38950082659721375, "learning_rate": 6.686064363032451e-05, "loss": 0.0255, "step": 42100 }, { "grad_norm": 0.08601085841655731, "learning_rate": 6.684507647382596e-05, "loss": 0.0147, "step": 42110 }, { "grad_norm": 0.27106809616088867, "learning_rate": 6.682950747517513e-05, "loss": 0.0282, "step": 42120 }, { "grad_norm": 0.23780235648155212, "learning_rate": 6.68139366360747e-05, "loss": 0.0226, "step": 42130 }, { "grad_norm": 0.44309306144714355, "learning_rate": 6.679836395822744e-05, "loss": 0.0209, "step": 42140 }, { "grad_norm": 0.14728914201259613, "learning_rate": 6.678278944333633e-05, "loss": 0.0273, "step": 42150 }, { "grad_norm": 0.33793872594833374, "learning_rate": 6.676721309310462e-05, "loss": 0.0202, "step": 42160 }, { "grad_norm": 0.19350560009479523, "learning_rate": 6.67516349092357e-05, "loss": 0.0195, "step": 42170 }, { "grad_norm": 0.11810142546892166, "learning_rate": 6.673605489343315e-05, "loss": 0.0125, "step": 42180 }, { "grad_norm": 0.27107539772987366, "learning_rate": 6.672047304740081e-05, "loss": 0.0217, "step": 42190 }, { "grad_norm": 0.19844646751880646, "learning_rate": 6.670488937284268e-05, "loss": 0.0184, "step": 42200 }, { "grad_norm": 0.3180698752403259, "learning_rate": 6.668930387146294e-05, "loss": 0.018, "step": 42210 }, { "grad_norm": 0.435545414686203, "learning_rate": 6.667371654496605e-05, "loss": 0.0259, "step": 42220 }, { "grad_norm": 0.13182808458805084, "learning_rate": 6.665812739505659e-05, "loss": 0.0233, "step": 42230 }, { "grad_norm": 0.35408592224121094, "learning_rate": 6.664253642343935e-05, "loss": 0.0253, "step": 42240 }, { "grad_norm": 0.1811935305595398, "learning_rate": 6.662694363181935e-05, "loss": 0.0152, "step": 42250 }, { "grad_norm": 0.3321867883205414, "learning_rate": 6.66113490219018e-05, "loss": 0.0276, "step": 42260 }, { "grad_norm": 0.11369109898805618, "learning_rate": 6.659575259539207e-05, "loss": 0.0156, "step": 42270 }, { "grad_norm": 0.22651880979537964, "learning_rate": 6.65801543539958e-05, "loss": 0.0203, "step": 42280 }, { "grad_norm": 0.3655405342578888, "learning_rate": 6.65645542994188e-05, "loss": 0.0191, "step": 42290 }, { "grad_norm": 0.19898532330989838, "learning_rate": 6.654895243336702e-05, "loss": 0.017, "step": 42300 }, { "grad_norm": 0.21595464646816254, "learning_rate": 6.65333487575467e-05, "loss": 0.0164, "step": 42310 }, { "grad_norm": 0.27204570174217224, "learning_rate": 6.651774327366422e-05, "loss": 0.0164, "step": 42320 }, { "grad_norm": 0.4664299190044403, "learning_rate": 6.650213598342613e-05, "loss": 0.0239, "step": 42330 }, { "grad_norm": 0.22334161400794983, "learning_rate": 6.648652688853931e-05, "loss": 0.0246, "step": 42340 }, { "grad_norm": 0.245527446269989, "learning_rate": 6.647091599071066e-05, "loss": 0.0244, "step": 42350 }, { "grad_norm": 0.30386441946029663, "learning_rate": 6.645530329164743e-05, "loss": 0.0224, "step": 42360 }, { "grad_norm": 0.17656461894512177, "learning_rate": 6.643968879305694e-05, "loss": 0.0149, "step": 42370 }, { "grad_norm": 0.1425361931324005, "learning_rate": 6.642407249664683e-05, "loss": 0.0202, "step": 42380 }, { "grad_norm": 0.14333468675613403, "learning_rate": 6.640845440412483e-05, "loss": 0.0275, "step": 42390 }, { "grad_norm": 0.26575177907943726, "learning_rate": 6.639283451719893e-05, "loss": 0.0182, "step": 42400 }, { "grad_norm": 0.11491065472364426, "learning_rate": 6.637721283757729e-05, "loss": 0.0215, "step": 42410 }, { "grad_norm": 0.2374432533979416, "learning_rate": 6.636158936696828e-05, "loss": 0.0208, "step": 42420 }, { "grad_norm": 0.4556857943534851, "learning_rate": 6.634596410708047e-05, "loss": 0.0243, "step": 42430 }, { "grad_norm": 0.25124162435531616, "learning_rate": 6.633033705962257e-05, "loss": 0.0229, "step": 42440 }, { "grad_norm": 0.16997648775577545, "learning_rate": 6.631470822630359e-05, "loss": 0.0195, "step": 42450 }, { "grad_norm": 0.16730771958827972, "learning_rate": 6.629907760883263e-05, "loss": 0.026, "step": 42460 }, { "grad_norm": 0.11038850247859955, "learning_rate": 6.628344520891905e-05, "loss": 0.0212, "step": 42470 }, { "grad_norm": 0.1462417095899582, "learning_rate": 6.62678110282724e-05, "loss": 0.0133, "step": 42480 }, { "grad_norm": 0.18033747375011444, "learning_rate": 6.625217506860237e-05, "loss": 0.0195, "step": 42490 }, { "grad_norm": 0.1702856421470642, "learning_rate": 6.623653733161892e-05, "loss": 0.0221, "step": 42500 }, { "grad_norm": 0.13666564226150513, "learning_rate": 6.622089781903217e-05, "loss": 0.0222, "step": 42510 }, { "grad_norm": 0.2795035243034363, "learning_rate": 6.620525653255241e-05, "loss": 0.0234, "step": 42520 }, { "grad_norm": 0.32278159260749817, "learning_rate": 6.618961347389015e-05, "loss": 0.0185, "step": 42530 }, { "grad_norm": 0.24585659801959991, "learning_rate": 6.617396864475613e-05, "loss": 0.0134, "step": 42540 }, { "grad_norm": 0.2606087327003479, "learning_rate": 6.61583220468612e-05, "loss": 0.0199, "step": 42550 }, { "grad_norm": 0.14020590484142303, "learning_rate": 6.614267368191645e-05, "loss": 0.0178, "step": 42560 }, { "grad_norm": 0.3012782335281372, "learning_rate": 6.61270235516332e-05, "loss": 0.018, "step": 42570 }, { "grad_norm": 0.11980891972780228, "learning_rate": 6.611137165772287e-05, "loss": 0.0109, "step": 42580 }, { "grad_norm": 0.3506942391395569, "learning_rate": 6.60957180018972e-05, "loss": 0.0229, "step": 42590 }, { "grad_norm": 0.3947935402393341, "learning_rate": 6.608006258586797e-05, "loss": 0.018, "step": 42600 }, { "grad_norm": 0.14649000763893127, "learning_rate": 6.606440541134731e-05, "loss": 0.0188, "step": 42610 }, { "grad_norm": 0.17580080032348633, "learning_rate": 6.60487464800474e-05, "loss": 0.0278, "step": 42620 }, { "grad_norm": 0.3609853684902191, "learning_rate": 6.603308579368071e-05, "loss": 0.0241, "step": 42630 }, { "grad_norm": 0.14207874238491058, "learning_rate": 6.601742335395987e-05, "loss": 0.0174, "step": 42640 }, { "grad_norm": 0.26677945256233215, "learning_rate": 6.600175916259769e-05, "loss": 0.0247, "step": 42650 }, { "grad_norm": 0.2836397588253021, "learning_rate": 6.598609322130718e-05, "loss": 0.0165, "step": 42660 }, { "grad_norm": 0.3621215224266052, "learning_rate": 6.597042553180155e-05, "loss": 0.018, "step": 42670 }, { "grad_norm": 0.2523939609527588, "learning_rate": 6.59547560957942e-05, "loss": 0.0176, "step": 42680 }, { "grad_norm": 0.16940151154994965, "learning_rate": 6.59390849149987e-05, "loss": 0.029, "step": 42690 }, { "grad_norm": 0.2458162158727646, "learning_rate": 6.592341199112886e-05, "loss": 0.0244, "step": 42700 }, { "grad_norm": 0.1753038614988327, "learning_rate": 6.590773732589861e-05, "loss": 0.027, "step": 42710 }, { "grad_norm": 0.46852853894233704, "learning_rate": 6.589206092102213e-05, "loss": 0.0347, "step": 42720 }, { "grad_norm": 0.14455489814281464, "learning_rate": 6.587638277821375e-05, "loss": 0.015, "step": 42730 }, { "grad_norm": 0.29272615909576416, "learning_rate": 6.586070289918803e-05, "loss": 0.0151, "step": 42740 }, { "grad_norm": 0.1973968744277954, "learning_rate": 6.584502128565968e-05, "loss": 0.02, "step": 42750 }, { "grad_norm": 0.26427987217903137, "learning_rate": 6.582933793934363e-05, "loss": 0.0218, "step": 42760 }, { "grad_norm": 0.18476036190986633, "learning_rate": 6.581365286195498e-05, "loss": 0.0198, "step": 42770 }, { "grad_norm": 0.15548259019851685, "learning_rate": 6.579796605520903e-05, "loss": 0.0251, "step": 42780 }, { "grad_norm": 0.13185936212539673, "learning_rate": 6.578227752082127e-05, "loss": 0.0168, "step": 42790 }, { "grad_norm": 0.30476197600364685, "learning_rate": 6.576658726050735e-05, "loss": 0.0217, "step": 42800 }, { "grad_norm": 0.10489495098590851, "learning_rate": 6.575089527598316e-05, "loss": 0.0171, "step": 42810 }, { "grad_norm": 0.3628699779510498, "learning_rate": 6.573520156896476e-05, "loss": 0.0241, "step": 42820 }, { "grad_norm": 0.21785275638103485, "learning_rate": 6.571950614116835e-05, "loss": 0.0242, "step": 42830 }, { "grad_norm": 0.2429940104484558, "learning_rate": 6.570380899431039e-05, "loss": 0.0226, "step": 42840 }, { "grad_norm": 0.18495303392410278, "learning_rate": 6.568811013010749e-05, "loss": 0.024, "step": 42850 }, { "grad_norm": 0.29834452271461487, "learning_rate": 6.567240955027645e-05, "loss": 0.027, "step": 42860 }, { "grad_norm": 0.27615389227867126, "learning_rate": 6.565670725653427e-05, "loss": 0.0172, "step": 42870 }, { "grad_norm": 0.2405189722776413, "learning_rate": 6.564100325059812e-05, "loss": 0.019, "step": 42880 }, { "grad_norm": 0.2411595582962036, "learning_rate": 6.562529753418538e-05, "loss": 0.0238, "step": 42890 }, { "grad_norm": 0.20766423642635345, "learning_rate": 6.56095901090136e-05, "loss": 0.0167, "step": 42900 }, { "grad_norm": 0.3695424497127533, "learning_rate": 6.55938809768005e-05, "loss": 0.015, "step": 42910 }, { "grad_norm": 0.16358299553394318, "learning_rate": 6.557817013926402e-05, "loss": 0.0186, "step": 42920 }, { "grad_norm": 1.011165738105774, "learning_rate": 6.556245759812227e-05, "loss": 0.0314, "step": 42930 }, { "grad_norm": 0.2464466542005539, "learning_rate": 6.554674335509357e-05, "loss": 0.0239, "step": 42940 }, { "grad_norm": 0.2872312068939209, "learning_rate": 6.553102741189638e-05, "loss": 0.0189, "step": 42950 }, { "grad_norm": 0.1064080223441124, "learning_rate": 6.55153097702494e-05, "loss": 0.0209, "step": 42960 }, { "grad_norm": 0.35858479142189026, "learning_rate": 6.549959043187146e-05, "loss": 0.0269, "step": 42970 }, { "grad_norm": 0.2001718431711197, "learning_rate": 6.54838693984816e-05, "loss": 0.0189, "step": 42980 }, { "grad_norm": 0.25608405470848083, "learning_rate": 6.546814667179909e-05, "loss": 0.024, "step": 42990 }, { "grad_norm": 0.3124220073223114, "learning_rate": 6.545242225354328e-05, "loss": 0.0162, "step": 43000 }, { "grad_norm": 0.3374066650867462, "learning_rate": 6.543669614543383e-05, "loss": 0.0244, "step": 43010 }, { "grad_norm": 0.2591395974159241, "learning_rate": 6.542096834919049e-05, "loss": 0.0197, "step": 43020 }, { "grad_norm": 0.2818484604358673, "learning_rate": 6.540523886653324e-05, "loss": 0.0216, "step": 43030 }, { "grad_norm": 0.2684948146343231, "learning_rate": 6.538950769918222e-05, "loss": 0.0273, "step": 43040 }, { "grad_norm": 0.21762606501579285, "learning_rate": 6.537377484885779e-05, "loss": 0.0162, "step": 43050 }, { "grad_norm": 0.20325614511966705, "learning_rate": 6.535804031728045e-05, "loss": 0.0179, "step": 43060 }, { "grad_norm": 0.27181434631347656, "learning_rate": 6.534230410617092e-05, "loss": 0.0235, "step": 43070 }, { "grad_norm": 0.22840732336044312, "learning_rate": 6.532656621725007e-05, "loss": 0.0225, "step": 43080 }, { "grad_norm": 0.31944987177848816, "learning_rate": 6.531082665223898e-05, "loss": 0.0263, "step": 43090 }, { "grad_norm": 0.24160265922546387, "learning_rate": 6.529508541285889e-05, "loss": 0.0193, "step": 43100 }, { "grad_norm": 0.2645743191242218, "learning_rate": 6.527934250083128e-05, "loss": 0.0241, "step": 43110 }, { "grad_norm": 0.3027959167957306, "learning_rate": 6.526359791787775e-05, "loss": 0.0219, "step": 43120 }, { "grad_norm": 0.21058443188667297, "learning_rate": 6.524785166572009e-05, "loss": 0.0217, "step": 43130 }, { "grad_norm": 0.2400328367948532, "learning_rate": 6.523210374608029e-05, "loss": 0.0225, "step": 43140 }, { "grad_norm": 0.28232771158218384, "learning_rate": 6.521635416068054e-05, "loss": 0.0251, "step": 43150 }, { "grad_norm": 0.23704411089420319, "learning_rate": 6.520060291124317e-05, "loss": 0.0339, "step": 43160 }, { "grad_norm": 0.2576378583908081, "learning_rate": 6.518484999949072e-05, "loss": 0.0204, "step": 43170 }, { "grad_norm": 0.204163059592247, "learning_rate": 6.516909542714591e-05, "loss": 0.0237, "step": 43180 }, { "grad_norm": 0.1616002321243286, "learning_rate": 6.515333919593161e-05, "loss": 0.017, "step": 43190 }, { "grad_norm": 0.13808032870292664, "learning_rate": 6.513758130757094e-05, "loss": 0.0206, "step": 43200 }, { "grad_norm": 0.28384068608283997, "learning_rate": 6.512182176378713e-05, "loss": 0.0171, "step": 43210 }, { "grad_norm": 0.08190298080444336, "learning_rate": 6.510606056630362e-05, "loss": 0.0143, "step": 43220 }, { "grad_norm": 0.3120896518230438, "learning_rate": 6.509029771684405e-05, "loss": 0.0204, "step": 43230 }, { "grad_norm": 0.10741990804672241, "learning_rate": 6.50745332171322e-05, "loss": 0.0211, "step": 43240 }, { "grad_norm": 0.1924784779548645, "learning_rate": 6.505876706889207e-05, "loss": 0.0213, "step": 43250 }, { "grad_norm": 0.3957315683364868, "learning_rate": 6.504299927384781e-05, "loss": 0.0269, "step": 43260 }, { "grad_norm": 0.29258275032043457, "learning_rate": 6.502722983372378e-05, "loss": 0.0174, "step": 43270 }, { "grad_norm": 0.32335105538368225, "learning_rate": 6.501145875024446e-05, "loss": 0.0174, "step": 43280 }, { "grad_norm": 0.12413196265697479, "learning_rate": 6.499568602513459e-05, "loss": 0.0225, "step": 43290 }, { "grad_norm": 0.2166275829076767, "learning_rate": 6.497991166011903e-05, "loss": 0.0237, "step": 43300 }, { "grad_norm": 0.30124521255493164, "learning_rate": 6.49641356569229e-05, "loss": 0.0306, "step": 43310 }, { "grad_norm": 0.17654642462730408, "learning_rate": 6.494835801727135e-05, "loss": 0.0197, "step": 43320 }, { "grad_norm": 0.30051493644714355, "learning_rate": 6.493257874288986e-05, "loss": 0.0215, "step": 43330 }, { "grad_norm": 0.23616518080234528, "learning_rate": 6.491679783550399e-05, "loss": 0.0225, "step": 43340 }, { "grad_norm": 0.2408989667892456, "learning_rate": 6.490101529683957e-05, "loss": 0.0208, "step": 43350 }, { "grad_norm": 0.40273815393447876, "learning_rate": 6.488523112862248e-05, "loss": 0.0194, "step": 43360 }, { "grad_norm": 0.18591386079788208, "learning_rate": 6.486944533257891e-05, "loss": 0.022, "step": 43370 }, { "grad_norm": 0.31429481506347656, "learning_rate": 6.485365791043517e-05, "loss": 0.0218, "step": 43380 }, { "grad_norm": 0.17662791907787323, "learning_rate": 6.483786886391772e-05, "loss": 0.0134, "step": 43390 }, { "grad_norm": 0.10675723850727081, "learning_rate": 6.482207819475323e-05, "loss": 0.0147, "step": 43400 }, { "grad_norm": 0.22928301990032196, "learning_rate": 6.480628590466857e-05, "loss": 0.0162, "step": 43410 }, { "grad_norm": 0.10262508690357208, "learning_rate": 6.479049199539074e-05, "loss": 0.0176, "step": 43420 }, { "grad_norm": 0.19148637354373932, "learning_rate": 6.477469646864693e-05, "loss": 0.0197, "step": 43430 }, { "grad_norm": 0.22430266439914703, "learning_rate": 6.475889932616454e-05, "loss": 0.0221, "step": 43440 }, { "grad_norm": 0.19983690977096558, "learning_rate": 6.474310056967111e-05, "loss": 0.0196, "step": 43450 }, { "grad_norm": 0.21572767198085785, "learning_rate": 6.472730020089437e-05, "loss": 0.0298, "step": 43460 }, { "grad_norm": 0.1982773244380951, "learning_rate": 6.471149822156222e-05, "loss": 0.0207, "step": 43470 }, { "grad_norm": 0.13052484393119812, "learning_rate": 6.469569463340275e-05, "loss": 0.0159, "step": 43480 }, { "grad_norm": 0.14087772369384766, "learning_rate": 6.467988943814418e-05, "loss": 0.0156, "step": 43490 }, { "grad_norm": 0.3105558156967163, "learning_rate": 6.4664082637515e-05, "loss": 0.0269, "step": 43500 }, { "grad_norm": 0.4833488464355469, "learning_rate": 6.464827423324377e-05, "loss": 0.0221, "step": 43510 }, { "grad_norm": 0.27592384815216064, "learning_rate": 6.463246422705931e-05, "loss": 0.014, "step": 43520 }, { "grad_norm": 0.15471747517585754, "learning_rate": 6.461665262069055e-05, "loss": 0.0232, "step": 43530 }, { "grad_norm": 0.14931367337703705, "learning_rate": 6.460083941586665e-05, "loss": 0.0151, "step": 43540 }, { "grad_norm": 0.31043264269828796, "learning_rate": 6.45850246143169e-05, "loss": 0.0185, "step": 43550 }, { "grad_norm": 0.24608023464679718, "learning_rate": 6.456920821777077e-05, "loss": 0.0222, "step": 43560 }, { "grad_norm": 0.3836282789707184, "learning_rate": 6.455339022795796e-05, "loss": 0.0187, "step": 43570 }, { "grad_norm": 0.14707310497760773, "learning_rate": 6.453757064660825e-05, "loss": 0.0255, "step": 43580 }, { "grad_norm": 0.22090718150138855, "learning_rate": 6.452174947545169e-05, "loss": 0.0122, "step": 43590 }, { "grad_norm": 0.2512708306312561, "learning_rate": 6.450592671621842e-05, "loss": 0.0196, "step": 43600 }, { "grad_norm": 0.4634067416191101, "learning_rate": 6.449010237063882e-05, "loss": 0.0186, "step": 43610 }, { "grad_norm": 0.2944187819957733, "learning_rate": 6.447427644044343e-05, "loss": 0.017, "step": 43620 }, { "grad_norm": 0.22585558891296387, "learning_rate": 6.44584489273629e-05, "loss": 0.0144, "step": 43630 }, { "grad_norm": 0.1414984166622162, "learning_rate": 6.444261983312817e-05, "loss": 0.0187, "step": 43640 }, { "grad_norm": 0.2932920455932617, "learning_rate": 6.442678915947023e-05, "loss": 0.0149, "step": 43650 }, { "grad_norm": 0.22056284546852112, "learning_rate": 6.441095690812032e-05, "loss": 0.0278, "step": 43660 }, { "grad_norm": 0.14329780638217926, "learning_rate": 6.439512308080982e-05, "loss": 0.0192, "step": 43670 }, { "grad_norm": 0.3591853678226471, "learning_rate": 6.43792876792703e-05, "loss": 0.0234, "step": 43680 }, { "grad_norm": 0.20875033736228943, "learning_rate": 6.43634507052335e-05, "loss": 0.0248, "step": 43690 }, { "grad_norm": 0.1863100677728653, "learning_rate": 6.434761216043135e-05, "loss": 0.0203, "step": 43700 }, { "grad_norm": 0.3050347566604614, "learning_rate": 6.433177204659589e-05, "loss": 0.0194, "step": 43710 }, { "grad_norm": 0.30628952383995056, "learning_rate": 6.431593036545938e-05, "loss": 0.0155, "step": 43720 }, { "grad_norm": 0.3426342010498047, "learning_rate": 6.430008711875426e-05, "loss": 0.0219, "step": 43730 }, { "grad_norm": 0.21765996515750885, "learning_rate": 6.42842423082131e-05, "loss": 0.0225, "step": 43740 }, { "grad_norm": 0.12604407966136932, "learning_rate": 6.426839593556868e-05, "loss": 0.0176, "step": 43750 }, { "grad_norm": 0.22155144810676575, "learning_rate": 6.425254800255392e-05, "loss": 0.0219, "step": 43760 }, { "grad_norm": 0.3427189290523529, "learning_rate": 6.423669851090196e-05, "loss": 0.0203, "step": 43770 }, { "grad_norm": 0.15076689422130585, "learning_rate": 6.422084746234604e-05, "loss": 0.0143, "step": 43780 }, { "grad_norm": 0.34236931800842285, "learning_rate": 6.420499485861965e-05, "loss": 0.0208, "step": 43790 }, { "grad_norm": 0.14540472626686096, "learning_rate": 6.418914070145636e-05, "loss": 0.0206, "step": 43800 }, { "grad_norm": 0.2624244689941406, "learning_rate": 6.417328499258996e-05, "loss": 0.0208, "step": 43810 }, { "grad_norm": 0.1870398372411728, "learning_rate": 6.415742773375446e-05, "loss": 0.016, "step": 43820 }, { "grad_norm": 0.17067979276180267, "learning_rate": 6.414156892668393e-05, "loss": 0.0198, "step": 43830 }, { "grad_norm": 0.2777029573917389, "learning_rate": 6.412570857311267e-05, "loss": 0.0231, "step": 43840 }, { "grad_norm": 0.5703243613243103, "learning_rate": 6.410984667477518e-05, "loss": 0.0269, "step": 43850 }, { "grad_norm": 0.5202464461326599, "learning_rate": 6.409398323340607e-05, "loss": 0.0232, "step": 43860 }, { "grad_norm": 0.1831647753715515, "learning_rate": 6.407811825074012e-05, "loss": 0.0239, "step": 43870 }, { "grad_norm": 0.24437180161476135, "learning_rate": 6.406225172851234e-05, "loss": 0.0253, "step": 43880 }, { "grad_norm": 0.1282067745923996, "learning_rate": 6.404638366845786e-05, "loss": 0.0269, "step": 43890 }, { "grad_norm": 0.18963906168937683, "learning_rate": 6.403051407231196e-05, "loss": 0.0166, "step": 43900 }, { "grad_norm": 0.3550989031791687, "learning_rate": 6.401464294181016e-05, "loss": 0.0209, "step": 43910 }, { "grad_norm": 0.273017555475235, "learning_rate": 6.399877027868808e-05, "loss": 0.0197, "step": 43920 }, { "grad_norm": 0.36441224813461304, "learning_rate": 6.39828960846815e-05, "loss": 0.0259, "step": 43930 }, { "grad_norm": 0.2831186056137085, "learning_rate": 6.396702036152644e-05, "loss": 0.0257, "step": 43940 }, { "grad_norm": 0.17887137830257416, "learning_rate": 6.395114311095905e-05, "loss": 0.0247, "step": 43950 }, { "grad_norm": 0.33742013573646545, "learning_rate": 6.393526433471559e-05, "loss": 0.0188, "step": 43960 }, { "grad_norm": 0.167007178068161, "learning_rate": 6.391938403453262e-05, "loss": 0.024, "step": 43970 }, { "grad_norm": 0.39417240023612976, "learning_rate": 6.390350221214671e-05, "loss": 0.0226, "step": 43980 }, { "grad_norm": 0.39330554008483887, "learning_rate": 6.38876188692947e-05, "loss": 0.0287, "step": 43990 }, { "grad_norm": 0.4425707757472992, "learning_rate": 6.387173400771358e-05, "loss": 0.0227, "step": 44000 }, { "grad_norm": 0.2961028516292572, "learning_rate": 6.385584762914047e-05, "loss": 0.0148, "step": 44010 }, { "grad_norm": 0.27074363827705383, "learning_rate": 6.38399597353127e-05, "loss": 0.0245, "step": 44020 }, { "grad_norm": 0.15762071311473846, "learning_rate": 6.382407032796775e-05, "loss": 0.0196, "step": 44030 }, { "grad_norm": 0.1447199434041977, "learning_rate": 6.380817940884325e-05, "loss": 0.0122, "step": 44040 }, { "grad_norm": 0.9685553908348083, "learning_rate": 6.379228697967702e-05, "loss": 0.0245, "step": 44050 }, { "grad_norm": 0.24791860580444336, "learning_rate": 6.3776393042207e-05, "loss": 0.0191, "step": 44060 }, { "grad_norm": 0.22419904172420502, "learning_rate": 6.376049759817137e-05, "loss": 0.0201, "step": 44070 }, { "grad_norm": 0.29946017265319824, "learning_rate": 6.37446006493084e-05, "loss": 0.0284, "step": 44080 }, { "grad_norm": 0.30701324343681335, "learning_rate": 6.372870219735656e-05, "loss": 0.0228, "step": 44090 }, { "grad_norm": 0.4221269488334656, "learning_rate": 6.371280224405452e-05, "loss": 0.0245, "step": 44100 }, { "grad_norm": 0.35433489084243774, "learning_rate": 6.369690079114101e-05, "loss": 0.0258, "step": 44110 }, { "grad_norm": 0.27382516860961914, "learning_rate": 6.368099784035504e-05, "loss": 0.0211, "step": 44120 }, { "grad_norm": 0.18108433485031128, "learning_rate": 6.366509339343572e-05, "loss": 0.022, "step": 44130 }, { "grad_norm": 0.22221152484416962, "learning_rate": 6.364918745212232e-05, "loss": 0.0185, "step": 44140 }, { "grad_norm": 0.17432597279548645, "learning_rate": 6.363328001815431e-05, "loss": 0.0209, "step": 44150 }, { "grad_norm": 0.2260240912437439, "learning_rate": 6.361737109327128e-05, "loss": 0.0193, "step": 44160 }, { "grad_norm": 0.20288248360157013, "learning_rate": 6.360146067921301e-05, "loss": 0.0167, "step": 44170 }, { "grad_norm": 0.20979303121566772, "learning_rate": 6.35855487777195e-05, "loss": 0.0168, "step": 44180 }, { "grad_norm": 0.11781778186559677, "learning_rate": 6.356963539053076e-05, "loss": 0.0345, "step": 44190 }, { "grad_norm": 0.17188525199890137, "learning_rate": 6.355372051938711e-05, "loss": 0.0223, "step": 44200 }, { "grad_norm": 0.3073370158672333, "learning_rate": 6.353780416602894e-05, "loss": 0.0174, "step": 44210 }, { "grad_norm": 0.24853911995887756, "learning_rate": 6.352188633219689e-05, "loss": 0.0142, "step": 44220 }, { "grad_norm": 0.19942717254161835, "learning_rate": 6.350596701963166e-05, "loss": 0.0181, "step": 44230 }, { "grad_norm": 0.14174681901931763, "learning_rate": 6.349004623007419e-05, "loss": 0.0167, "step": 44240 }, { "grad_norm": 0.4425675570964813, "learning_rate": 6.347412396526555e-05, "loss": 0.0235, "step": 44250 }, { "grad_norm": 0.18677228689193726, "learning_rate": 6.345820022694696e-05, "loss": 0.0176, "step": 44260 }, { "grad_norm": 0.34300994873046875, "learning_rate": 6.344227501685984e-05, "loss": 0.0198, "step": 44270 }, { "grad_norm": 0.3592073321342468, "learning_rate": 6.342634833674572e-05, "loss": 0.0237, "step": 44280 }, { "grad_norm": 0.269977331161499, "learning_rate": 6.341042018834635e-05, "loss": 0.0265, "step": 44290 }, { "grad_norm": 0.2516328990459442, "learning_rate": 6.339449057340359e-05, "loss": 0.0163, "step": 44300 }, { "grad_norm": 0.0975441038608551, "learning_rate": 6.337855949365945e-05, "loss": 0.0117, "step": 44310 }, { "grad_norm": 0.11547472327947617, "learning_rate": 6.336262695085619e-05, "loss": 0.0278, "step": 44320 }, { "grad_norm": 0.37527331709861755, "learning_rate": 6.334669294673612e-05, "loss": 0.0304, "step": 44330 }, { "grad_norm": 0.31525716185569763, "learning_rate": 6.333075748304179e-05, "loss": 0.0244, "step": 44340 }, { "grad_norm": 0.37206608057022095, "learning_rate": 6.331482056151585e-05, "loss": 0.0225, "step": 44350 }, { "grad_norm": 0.240166574716568, "learning_rate": 6.329888218390117e-05, "loss": 0.0235, "step": 44360 }, { "grad_norm": 0.3111308217048645, "learning_rate": 6.328294235194072e-05, "loss": 0.016, "step": 44370 }, { "grad_norm": 0.3382244110107422, "learning_rate": 6.326700106737765e-05, "loss": 0.0235, "step": 44380 }, { "grad_norm": 0.20423419773578644, "learning_rate": 6.325105833195531e-05, "loss": 0.0167, "step": 44390 }, { "grad_norm": 0.3574983477592468, "learning_rate": 6.323511414741715e-05, "loss": 0.0279, "step": 44400 }, { "grad_norm": 0.19970883429050446, "learning_rate": 6.321916851550678e-05, "loss": 0.0209, "step": 44410 }, { "grad_norm": 0.318096786737442, "learning_rate": 6.320322143796806e-05, "loss": 0.0214, "step": 44420 }, { "grad_norm": 0.22180578112602234, "learning_rate": 6.318727291654488e-05, "loss": 0.0177, "step": 44430 }, { "grad_norm": 0.2924504578113556, "learning_rate": 6.317132295298134e-05, "loss": 0.0264, "step": 44440 }, { "grad_norm": 0.2950877547264099, "learning_rate": 6.315537154902173e-05, "loss": 0.0159, "step": 44450 }, { "grad_norm": 0.20927579700946808, "learning_rate": 6.313941870641048e-05, "loss": 0.0204, "step": 44460 }, { "grad_norm": 0.21408146619796753, "learning_rate": 6.312346442689214e-05, "loss": 0.0245, "step": 44470 }, { "grad_norm": 0.2356046587228775, "learning_rate": 6.310750871221147e-05, "loss": 0.0135, "step": 44480 }, { "grad_norm": 0.22129814326763153, "learning_rate": 6.309155156411335e-05, "loss": 0.0188, "step": 44490 }, { "grad_norm": 0.1487596482038498, "learning_rate": 6.307559298434284e-05, "loss": 0.0274, "step": 44500 }, { "grad_norm": 0.10179734975099564, "learning_rate": 6.305963297464512e-05, "loss": 0.022, "step": 44510 }, { "grad_norm": 0.18763139843940735, "learning_rate": 6.304367153676561e-05, "loss": 0.0209, "step": 44520 }, { "grad_norm": 0.35528284311294556, "learning_rate": 6.302770867244978e-05, "loss": 0.0229, "step": 44530 }, { "grad_norm": 0.11394840478897095, "learning_rate": 6.301174438344328e-05, "loss": 0.019, "step": 44540 }, { "grad_norm": 0.41846033930778503, "learning_rate": 6.299577867149202e-05, "loss": 0.0213, "step": 44550 }, { "grad_norm": 0.45761653780937195, "learning_rate": 6.297981153834192e-05, "loss": 0.0258, "step": 44560 }, { "grad_norm": 0.23835869133472443, "learning_rate": 6.296384298573916e-05, "loss": 0.0369, "step": 44570 }, { "grad_norm": 0.20794633030891418, "learning_rate": 6.294787301543001e-05, "loss": 0.0434, "step": 44580 }, { "grad_norm": 0.5266525745391846, "learning_rate": 6.293190162916095e-05, "loss": 0.0301, "step": 44590 }, { "grad_norm": 0.3861885070800781, "learning_rate": 6.291592882867855e-05, "loss": 0.0276, "step": 44600 }, { "grad_norm": 0.30696621537208557, "learning_rate": 6.28999546157296e-05, "loss": 0.0194, "step": 44610 }, { "grad_norm": 0.1847144514322281, "learning_rate": 6.288397899206102e-05, "loss": 0.0139, "step": 44620 }, { "grad_norm": 0.3309484124183655, "learning_rate": 6.286800195941984e-05, "loss": 0.0257, "step": 44630 }, { "grad_norm": 0.2441725879907608, "learning_rate": 6.285202351955334e-05, "loss": 0.0209, "step": 44640 }, { "grad_norm": 0.1328764110803604, "learning_rate": 6.283604367420887e-05, "loss": 0.0138, "step": 44650 }, { "grad_norm": 0.5269983410835266, "learning_rate": 6.282006242513394e-05, "loss": 0.0212, "step": 44660 }, { "grad_norm": 0.1837935447692871, "learning_rate": 6.280407977407628e-05, "loss": 0.0209, "step": 44670 }, { "grad_norm": 0.32890161871910095, "learning_rate": 6.27880957227837e-05, "loss": 0.0233, "step": 44680 }, { "grad_norm": 0.23760908842086792, "learning_rate": 6.27721102730042e-05, "loss": 0.0168, "step": 44690 }, { "grad_norm": 0.30480143427848816, "learning_rate": 6.275612342648592e-05, "loss": 0.0271, "step": 44700 }, { "grad_norm": 0.22815248370170593, "learning_rate": 6.274013518497716e-05, "loss": 0.0164, "step": 44710 }, { "grad_norm": 0.21604354679584503, "learning_rate": 6.272414555022636e-05, "loss": 0.0167, "step": 44720 }, { "grad_norm": 0.14302025735378265, "learning_rate": 6.270815452398215e-05, "loss": 0.0187, "step": 44730 }, { "grad_norm": 0.16703979671001434, "learning_rate": 6.269216210799326e-05, "loss": 0.0194, "step": 44740 }, { "grad_norm": 0.2380751520395279, "learning_rate": 6.26761683040086e-05, "loss": 0.0407, "step": 44750 }, { "grad_norm": 0.1618916094303131, "learning_rate": 6.266017311377723e-05, "loss": 0.025, "step": 44760 }, { "grad_norm": 0.30370089411735535, "learning_rate": 6.264417653904839e-05, "loss": 0.0306, "step": 44770 }, { "grad_norm": 0.1946887969970703, "learning_rate": 6.262817858157139e-05, "loss": 0.0182, "step": 44780 }, { "grad_norm": 0.21756143867969513, "learning_rate": 6.261217924309576e-05, "loss": 0.0257, "step": 44790 }, { "grad_norm": 0.19339336454868317, "learning_rate": 6.259617852537118e-05, "loss": 0.0175, "step": 44800 }, { "grad_norm": 0.22996464371681213, "learning_rate": 6.258017643014747e-05, "loss": 0.0235, "step": 44810 }, { "grad_norm": 0.2668381333351135, "learning_rate": 6.256417295917456e-05, "loss": 0.0222, "step": 44820 }, { "grad_norm": 0.17459623515605927, "learning_rate": 6.254816811420258e-05, "loss": 0.0249, "step": 44830 }, { "grad_norm": 0.14837276935577393, "learning_rate": 6.253216189698183e-05, "loss": 0.012, "step": 44840 }, { "grad_norm": 0.166948601603508, "learning_rate": 6.251615430926267e-05, "loss": 0.0151, "step": 44850 }, { "grad_norm": 0.26856619119644165, "learning_rate": 6.25001453527957e-05, "loss": 0.0193, "step": 44860 }, { "grad_norm": 0.2511186897754669, "learning_rate": 6.248413502933164e-05, "loss": 0.0221, "step": 44870 }, { "grad_norm": 0.23703287541866302, "learning_rate": 6.246812334062133e-05, "loss": 0.0116, "step": 44880 }, { "grad_norm": 0.2085273414850235, "learning_rate": 6.245211028841579e-05, "loss": 0.0178, "step": 44890 }, { "grad_norm": 0.3253522217273712, "learning_rate": 6.24360958744662e-05, "loss": 0.0211, "step": 44900 }, { "grad_norm": 0.2408585399389267, "learning_rate": 6.242008010052387e-05, "loss": 0.0197, "step": 44910 }, { "grad_norm": 0.337673544883728, "learning_rate": 6.240406296834024e-05, "loss": 0.0289, "step": 44920 }, { "grad_norm": 0.12331569939851761, "learning_rate": 6.238804447966694e-05, "loss": 0.0157, "step": 44930 }, { "grad_norm": 0.14658890664577484, "learning_rate": 6.237202463625573e-05, "loss": 0.0202, "step": 44940 }, { "grad_norm": 0.7515140175819397, "learning_rate": 6.235600343985848e-05, "loss": 0.0279, "step": 44950 }, { "grad_norm": 0.2556953728199005, "learning_rate": 6.233998089222729e-05, "loss": 0.0242, "step": 44960 }, { "grad_norm": 0.29677754640579224, "learning_rate": 6.232395699511433e-05, "loss": 0.0184, "step": 44970 }, { "grad_norm": 0.24555884301662445, "learning_rate": 6.230793175027199e-05, "loss": 0.0287, "step": 44980 }, { "grad_norm": 0.16661474108695984, "learning_rate": 6.22919051594527e-05, "loss": 0.016, "step": 44990 }, { "grad_norm": 0.17845870554447174, "learning_rate": 6.227587722440917e-05, "loss": 0.0222, "step": 45000 }, { "grad_norm": 0.25131475925445557, "learning_rate": 6.225984794689414e-05, "loss": 0.0214, "step": 45010 }, { "grad_norm": 0.2070632427930832, "learning_rate": 6.22438173286606e-05, "loss": 0.0296, "step": 45020 }, { "grad_norm": 0.1907891482114792, "learning_rate": 6.222778537146159e-05, "loss": 0.0187, "step": 45030 }, { "grad_norm": 0.11913293600082397, "learning_rate": 6.221175207705037e-05, "loss": 0.0205, "step": 45040 }, { "grad_norm": 0.17734679579734802, "learning_rate": 6.219571744718027e-05, "loss": 0.0211, "step": 45050 }, { "grad_norm": 0.20291446149349213, "learning_rate": 6.217968148360487e-05, "loss": 0.0201, "step": 45060 }, { "grad_norm": 0.21485118567943573, "learning_rate": 6.216364418807782e-05, "loss": 0.0116, "step": 45070 }, { "grad_norm": 0.21959882974624634, "learning_rate": 6.214760556235292e-05, "loss": 0.0154, "step": 45080 }, { "grad_norm": 0.345445841550827, "learning_rate": 6.213156560818414e-05, "loss": 0.018, "step": 45090 }, { "grad_norm": 0.1953604519367218, "learning_rate": 6.211552432732559e-05, "loss": 0.0165, "step": 45100 }, { "grad_norm": 0.10866348445415497, "learning_rate": 6.20994817215315e-05, "loss": 0.0168, "step": 45110 }, { "grad_norm": 0.17739084362983704, "learning_rate": 6.208343779255628e-05, "loss": 0.0148, "step": 45120 }, { "grad_norm": 0.34130915999412537, "learning_rate": 6.206739254215449e-05, "loss": 0.021, "step": 45130 }, { "grad_norm": 0.3245471119880676, "learning_rate": 6.205134597208077e-05, "loss": 0.018, "step": 45140 }, { "grad_norm": 0.16897867619991302, "learning_rate": 6.203529808409e-05, "loss": 0.0175, "step": 45150 }, { "grad_norm": 0.19319182634353638, "learning_rate": 6.201924887993712e-05, "loss": 0.0196, "step": 45160 }, { "grad_norm": 0.17371195554733276, "learning_rate": 6.200319836137725e-05, "loss": 0.0255, "step": 45170 }, { "grad_norm": 0.15279600024223328, "learning_rate": 6.198714653016565e-05, "loss": 0.0179, "step": 45180 }, { "grad_norm": 0.26725664734840393, "learning_rate": 6.197109338805774e-05, "loss": 0.0232, "step": 45190 }, { "grad_norm": 0.2781578302383423, "learning_rate": 6.195503893680903e-05, "loss": 0.0293, "step": 45200 }, { "grad_norm": 0.22718511521816254, "learning_rate": 6.193898317817524e-05, "loss": 0.0194, "step": 45210 }, { "grad_norm": 0.140061154961586, "learning_rate": 6.192292611391221e-05, "loss": 0.0169, "step": 45220 }, { "grad_norm": 0.3198590874671936, "learning_rate": 6.190686774577591e-05, "loss": 0.0222, "step": 45230 }, { "grad_norm": 0.15762677788734436, "learning_rate": 6.189080807552245e-05, "loss": 0.0199, "step": 45240 }, { "grad_norm": 0.15344758331775665, "learning_rate": 6.187474710490809e-05, "loss": 0.0229, "step": 45250 }, { "grad_norm": 0.32672998309135437, "learning_rate": 6.185868483568926e-05, "loss": 0.0215, "step": 45260 }, { "grad_norm": 0.15535248816013336, "learning_rate": 6.184262126962245e-05, "loss": 0.023, "step": 45270 }, { "grad_norm": 0.12893617153167725, "learning_rate": 6.182655640846442e-05, "loss": 0.02, "step": 45280 }, { "grad_norm": 0.2883732318878174, "learning_rate": 6.181049025397196e-05, "loss": 0.0202, "step": 45290 }, { "grad_norm": 0.27920621633529663, "learning_rate": 6.179442280790202e-05, "loss": 0.0293, "step": 45300 }, { "grad_norm": 0.21329958736896515, "learning_rate": 6.177835407201174e-05, "loss": 0.0203, "step": 45310 }, { "grad_norm": 0.35339662432670593, "learning_rate": 6.176228404805839e-05, "loss": 0.0209, "step": 45320 }, { "grad_norm": 0.28941401839256287, "learning_rate": 6.174621273779932e-05, "loss": 0.0215, "step": 45330 }, { "grad_norm": 0.30531883239746094, "learning_rate": 6.17301401429921e-05, "loss": 0.0195, "step": 45340 }, { "grad_norm": 0.1992606371641159, "learning_rate": 6.171406626539436e-05, "loss": 0.0136, "step": 45350 }, { "grad_norm": 0.4755226671695709, "learning_rate": 6.169799110676398e-05, "loss": 0.0245, "step": 45360 }, { "grad_norm": 0.23841159045696259, "learning_rate": 6.168191466885885e-05, "loss": 0.0214, "step": 45370 }, { "grad_norm": 0.21916009485721588, "learning_rate": 6.166583695343711e-05, "loss": 0.0228, "step": 45380 }, { "grad_norm": 0.1440727263689041, "learning_rate": 6.164975796225698e-05, "loss": 0.0215, "step": 45390 }, { "grad_norm": 0.13911189138889313, "learning_rate": 6.163367769707683e-05, "loss": 0.0212, "step": 45400 }, { "grad_norm": 0.1509319692850113, "learning_rate": 6.161759615965519e-05, "loss": 0.0241, "step": 45410 }, { "grad_norm": 0.20346137881278992, "learning_rate": 6.16015133517507e-05, "loss": 0.0205, "step": 45420 }, { "grad_norm": 0.27495187520980835, "learning_rate": 6.158542927512214e-05, "loss": 0.0193, "step": 45430 }, { "grad_norm": 0.1746865212917328, "learning_rate": 6.156934393152846e-05, "loss": 0.0184, "step": 45440 }, { "grad_norm": 0.35445964336395264, "learning_rate": 6.15532573227287e-05, "loss": 0.0257, "step": 45450 }, { "grad_norm": 0.2474350929260254, "learning_rate": 6.153716945048212e-05, "loss": 0.0257, "step": 45460 }, { "grad_norm": 0.20087815821170807, "learning_rate": 6.152108031654802e-05, "loss": 0.0234, "step": 45470 }, { "grad_norm": 0.2590954005718231, "learning_rate": 6.15049899226859e-05, "loss": 0.0242, "step": 45480 }, { "grad_norm": 0.2969464063644409, "learning_rate": 6.148889827065537e-05, "loss": 0.02, "step": 45490 }, { "grad_norm": 0.25768405199050903, "learning_rate": 6.147280536221622e-05, "loss": 0.0212, "step": 45500 }, { "grad_norm": 0.12262208759784698, "learning_rate": 6.145671119912832e-05, "loss": 0.0183, "step": 45510 }, { "grad_norm": 0.20193824172019958, "learning_rate": 6.144061578315169e-05, "loss": 0.0191, "step": 45520 }, { "grad_norm": 0.1819140762090683, "learning_rate": 6.142451911604654e-05, "loss": 0.016, "step": 45530 }, { "grad_norm": 0.1833905279636383, "learning_rate": 6.140842119957315e-05, "loss": 0.0147, "step": 45540 }, { "grad_norm": 0.21476416289806366, "learning_rate": 6.139232203549197e-05, "loss": 0.0187, "step": 45550 }, { "grad_norm": 0.09534615278244019, "learning_rate": 6.137622162556357e-05, "loss": 0.0211, "step": 45560 }, { "grad_norm": 0.31204909086227417, "learning_rate": 6.13601199715487e-05, "loss": 0.0156, "step": 45570 }, { "grad_norm": 0.4036891758441925, "learning_rate": 6.134401707520816e-05, "loss": 0.0199, "step": 45580 }, { "grad_norm": 0.15658284723758698, "learning_rate": 6.132791293830298e-05, "loss": 0.0122, "step": 45590 }, { "grad_norm": 0.27704986929893494, "learning_rate": 6.131180756259428e-05, "loss": 0.0234, "step": 45600 }, { "grad_norm": 0.28423285484313965, "learning_rate": 6.129570094984331e-05, "loss": 0.0186, "step": 45610 }, { "grad_norm": 0.19961658120155334, "learning_rate": 6.127959310181145e-05, "loss": 0.0256, "step": 45620 }, { "grad_norm": 0.18421034514904022, "learning_rate": 6.126348402026026e-05, "loss": 0.0184, "step": 45630 }, { "grad_norm": 0.38734251260757446, "learning_rate": 6.12473737069514e-05, "loss": 0.0225, "step": 45640 }, { "grad_norm": 0.3549957871437073, "learning_rate": 6.123126216364665e-05, "loss": 0.0253, "step": 45650 }, { "grad_norm": 0.266627699136734, "learning_rate": 6.121514939210797e-05, "loss": 0.023, "step": 45660 }, { "grad_norm": 0.3697468638420105, "learning_rate": 6.119903539409741e-05, "loss": 0.0145, "step": 45670 }, { "grad_norm": 0.224387064576149, "learning_rate": 6.118292017137716e-05, "loss": 0.0229, "step": 45680 }, { "grad_norm": 0.14108403027057648, "learning_rate": 6.116680372570959e-05, "loss": 0.0216, "step": 45690 }, { "grad_norm": 0.3205066919326782, "learning_rate": 6.115068605885713e-05, "loss": 0.022, "step": 45700 }, { "grad_norm": 0.1620820015668869, "learning_rate": 6.113456717258243e-05, "loss": 0.0229, "step": 45710 }, { "grad_norm": 0.2561241686344147, "learning_rate": 6.11184470686482e-05, "loss": 0.0184, "step": 45720 }, { "grad_norm": 0.1612798571586609, "learning_rate": 6.110232574881731e-05, "loss": 0.0191, "step": 45730 }, { "grad_norm": 0.19415071606636047, "learning_rate": 6.108620321485277e-05, "loss": 0.0165, "step": 45740 }, { "grad_norm": 0.2474934607744217, "learning_rate": 6.107007946851773e-05, "loss": 0.0209, "step": 45750 }, { "grad_norm": 0.18540355563163757, "learning_rate": 6.105395451157542e-05, "loss": 0.0184, "step": 45760 }, { "grad_norm": 0.23825576901435852, "learning_rate": 6.103782834578928e-05, "loss": 0.0205, "step": 45770 }, { "grad_norm": 0.17802579700946808, "learning_rate": 6.102170097292281e-05, "loss": 0.0286, "step": 45780 }, { "grad_norm": 0.29911884665489197, "learning_rate": 6.1005572394739716e-05, "loss": 0.0271, "step": 45790 }, { "grad_norm": 0.2796507179737091, "learning_rate": 6.0989442613003765e-05, "loss": 0.0177, "step": 45800 }, { "grad_norm": 0.3290281891822815, "learning_rate": 6.097331162947888e-05, "loss": 0.0284, "step": 45810 }, { "grad_norm": 0.3961332142353058, "learning_rate": 6.095717944592914e-05, "loss": 0.0254, "step": 45820 }, { "grad_norm": 0.22158755362033844, "learning_rate": 6.094104606411873e-05, "loss": 0.0255, "step": 45830 }, { "grad_norm": 0.20196950435638428, "learning_rate": 6.0924911485811966e-05, "loss": 0.0284, "step": 45840 }, { "grad_norm": 0.2076733112335205, "learning_rate": 6.090877571277331e-05, "loss": 0.0226, "step": 45850 }, { "grad_norm": 0.4480821490287781, "learning_rate": 6.089263874676736e-05, "loss": 0.0254, "step": 45860 }, { "grad_norm": 0.5993524193763733, "learning_rate": 6.0876500589558796e-05, "loss": 0.0276, "step": 45870 }, { "grad_norm": 0.15238887071609497, "learning_rate": 6.086036124291248e-05, "loss": 0.0181, "step": 45880 }, { "grad_norm": 0.232942134141922, "learning_rate": 6.084422070859339e-05, "loss": 0.0385, "step": 45890 }, { "grad_norm": 0.2385639101266861, "learning_rate": 6.082807898836663e-05, "loss": 0.0191, "step": 45900 }, { "grad_norm": 0.31097468733787537, "learning_rate": 6.081193608399742e-05, "loss": 0.0169, "step": 45910 }, { "grad_norm": 0.2229865938425064, "learning_rate": 6.0795791997251164e-05, "loss": 0.0173, "step": 45920 }, { "grad_norm": 0.2699875831604004, "learning_rate": 6.0779646729893294e-05, "loss": 0.0207, "step": 45930 }, { "grad_norm": 0.2142493724822998, "learning_rate": 6.0763500283689476e-05, "loss": 0.0173, "step": 45940 }, { "grad_norm": 0.13770057260990143, "learning_rate": 6.0747352660405455e-05, "loss": 0.0235, "step": 45950 }, { "grad_norm": 0.2093954235315323, "learning_rate": 6.073120386180709e-05, "loss": 0.0151, "step": 45960 }, { "grad_norm": 0.12779337167739868, "learning_rate": 6.0715053889660425e-05, "loss": 0.0186, "step": 45970 }, { "grad_norm": 0.306098997592926, "learning_rate": 6.069890274573157e-05, "loss": 0.0184, "step": 45980 }, { "grad_norm": 0.26765191555023193, "learning_rate": 6.068275043178679e-05, "loss": 0.0182, "step": 45990 }, { "grad_norm": 0.25712937116622925, "learning_rate": 6.066659694959248e-05, "loss": 0.0226, "step": 46000 }, { "grad_norm": 0.16432173550128937, "learning_rate": 6.0650442300915176e-05, "loss": 0.0132, "step": 46010 }, { "grad_norm": 0.14163970947265625, "learning_rate": 6.063428648752152e-05, "loss": 0.0171, "step": 46020 }, { "grad_norm": 0.5324731469154358, "learning_rate": 6.0618129511178266e-05, "loss": 0.021, "step": 46030 }, { "grad_norm": 0.20208105444908142, "learning_rate": 6.060197137365234e-05, "loss": 0.0176, "step": 46040 }, { "grad_norm": 0.11542562395334244, "learning_rate": 6.058581207671077e-05, "loss": 0.0211, "step": 46050 }, { "grad_norm": 0.16355566680431366, "learning_rate": 6.056965162212072e-05, "loss": 0.021, "step": 46060 }, { "grad_norm": 0.14011700451374054, "learning_rate": 6.055349001164943e-05, "loss": 0.0252, "step": 46070 }, { "grad_norm": 0.1559605896472931, "learning_rate": 6.0537327247064347e-05, "loss": 0.0224, "step": 46080 }, { "grad_norm": 0.1784733533859253, "learning_rate": 6.0521163330133014e-05, "loss": 0.0227, "step": 46090 }, { "grad_norm": 0.15697714686393738, "learning_rate": 6.050499826262306e-05, "loss": 0.0233, "step": 46100 }, { "grad_norm": 0.10082148015499115, "learning_rate": 6.0488832046302294e-05, "loss": 0.0211, "step": 46110 }, { "grad_norm": 0.203911691904068, "learning_rate": 6.0472664682938626e-05, "loss": 0.0134, "step": 46120 }, { "grad_norm": 0.19661802053451538, "learning_rate": 6.045649617430009e-05, "loss": 0.0208, "step": 46130 }, { "grad_norm": 0.14819391071796417, "learning_rate": 6.0440326522154866e-05, "loss": 0.0179, "step": 46140 }, { "grad_norm": 0.13765504956245422, "learning_rate": 6.0424155728271224e-05, "loss": 0.0183, "step": 46150 }, { "grad_norm": 0.20464429259300232, "learning_rate": 6.040798379441758e-05, "loss": 0.0167, "step": 46160 }, { "grad_norm": 0.24472129344940186, "learning_rate": 6.0391810722362485e-05, "loss": 0.0265, "step": 46170 }, { "grad_norm": 0.16346682608127594, "learning_rate": 6.037563651387458e-05, "loss": 0.0189, "step": 46180 }, { "grad_norm": 0.28502726554870605, "learning_rate": 6.0359461170722666e-05, "loss": 0.0215, "step": 46190 }, { "grad_norm": 0.24318942427635193, "learning_rate": 6.034328469467566e-05, "loss": 0.0182, "step": 46200 }, { "grad_norm": 0.3055498003959656, "learning_rate": 6.0327107087502596e-05, "loss": 0.0294, "step": 46210 }, { "grad_norm": 0.15169070661067963, "learning_rate": 6.031092835097262e-05, "loss": 0.0176, "step": 46220 }, { "grad_norm": 0.17439962923526764, "learning_rate": 6.0294748486855024e-05, "loss": 0.0178, "step": 46230 }, { "grad_norm": 0.20065641403198242, "learning_rate": 6.0278567496919216e-05, "loss": 0.023, "step": 46240 }, { "grad_norm": 0.41023626923561096, "learning_rate": 6.026238538293472e-05, "loss": 0.0235, "step": 46250 }, { "grad_norm": 0.2408868819475174, "learning_rate": 6.024620214667118e-05, "loss": 0.0216, "step": 46260 }, { "grad_norm": 0.2591613233089447, "learning_rate": 6.0230017789898384e-05, "loss": 0.0118, "step": 46270 }, { "grad_norm": 0.09304554015398026, "learning_rate": 6.021383231438622e-05, "loss": 0.0241, "step": 46280 }, { "grad_norm": 0.2345569133758545, "learning_rate": 6.0197645721904704e-05, "loss": 0.0174, "step": 46290 }, { "grad_norm": 0.29075050354003906, "learning_rate": 6.0181458014224e-05, "loss": 0.0152, "step": 46300 }, { "grad_norm": 0.16133509576320648, "learning_rate": 6.016526919311435e-05, "loss": 0.0111, "step": 46310 }, { "grad_norm": 0.3943917751312256, "learning_rate": 6.014907926034613e-05, "loss": 0.016, "step": 46320 }, { "grad_norm": 0.1466837227344513, "learning_rate": 6.0132888217689875e-05, "loss": 0.0161, "step": 46330 }, { "grad_norm": 0.3425769805908203, "learning_rate": 6.01166960669162e-05, "loss": 0.0238, "step": 46340 }, { "grad_norm": 0.18022102117538452, "learning_rate": 6.0100502809795845e-05, "loss": 0.0186, "step": 46350 }, { "grad_norm": 0.17288453876972198, "learning_rate": 6.00843084480997e-05, "loss": 0.0194, "step": 46360 }, { "grad_norm": 0.3760455846786499, "learning_rate": 6.0068112983598736e-05, "loss": 0.0221, "step": 46370 }, { "grad_norm": 0.3851189613342285, "learning_rate": 6.005191641806407e-05, "loss": 0.0232, "step": 46380 }, { "grad_norm": 0.40901297330856323, "learning_rate": 6.003571875326694e-05, "loss": 0.0185, "step": 46390 }, { "grad_norm": 0.11513715982437134, "learning_rate": 6.00195199909787e-05, "loss": 0.0138, "step": 46400 }, { "grad_norm": 0.24675242602825165, "learning_rate": 6.000332013297082e-05, "loss": 0.0189, "step": 46410 }, { "grad_norm": 0.15854690968990326, "learning_rate": 5.998711918101487e-05, "loss": 0.017, "step": 46420 }, { "grad_norm": 0.3728768229484558, "learning_rate": 5.997091713688261e-05, "loss": 0.0176, "step": 46430 }, { "grad_norm": 0.39028674364089966, "learning_rate": 5.9954714002345836e-05, "loss": 0.0165, "step": 46440 }, { "grad_norm": 0.2852288484573364, "learning_rate": 5.993850977917649e-05, "loss": 0.0168, "step": 46450 }, { "grad_norm": 0.3561843931674957, "learning_rate": 5.992230446914667e-05, "loss": 0.023, "step": 46460 }, { "grad_norm": 0.2713783383369446, "learning_rate": 5.990609807402855e-05, "loss": 0.0225, "step": 46470 }, { "grad_norm": 0.176371231675148, "learning_rate": 5.988989059559443e-05, "loss": 0.0201, "step": 46480 }, { "grad_norm": 0.20651382207870483, "learning_rate": 5.987368203561675e-05, "loss": 0.0227, "step": 46490 }, { "grad_norm": 0.30009812116622925, "learning_rate": 5.9857472395868055e-05, "loss": 0.0156, "step": 46500 }, { "grad_norm": 0.2056763470172882, "learning_rate": 5.9841261678120983e-05, "loss": 0.0215, "step": 46510 }, { "grad_norm": 0.14665405452251434, "learning_rate": 5.982504988414834e-05, "loss": 0.0155, "step": 46520 }, { "grad_norm": 0.3998331129550934, "learning_rate": 5.9808837015723015e-05, "loss": 0.0364, "step": 46530 }, { "grad_norm": 0.16350795328617096, "learning_rate": 5.9792623074618016e-05, "loss": 0.0187, "step": 46540 }, { "grad_norm": 0.3632943332195282, "learning_rate": 5.977640806260648e-05, "loss": 0.0267, "step": 46550 }, { "grad_norm": 0.17342498898506165, "learning_rate": 5.976019198146166e-05, "loss": 0.0187, "step": 46560 }, { "grad_norm": 0.19884903728961945, "learning_rate": 5.974397483295692e-05, "loss": 0.0176, "step": 46570 }, { "grad_norm": 0.24184802174568176, "learning_rate": 5.972775661886575e-05, "loss": 0.0244, "step": 46580 }, { "grad_norm": 0.2817915380001068, "learning_rate": 5.971153734096173e-05, "loss": 0.0146, "step": 46590 }, { "grad_norm": 0.37942615151405334, "learning_rate": 5.969531700101859e-05, "loss": 0.0161, "step": 46600 }, { "grad_norm": 0.3399278521537781, "learning_rate": 5.9679095600810155e-05, "loss": 0.0214, "step": 46610 }, { "grad_norm": 0.2361333817243576, "learning_rate": 5.9662873142110384e-05, "loss": 0.0208, "step": 46620 }, { "grad_norm": 0.16911335289478302, "learning_rate": 5.964664962669333e-05, "loss": 0.0163, "step": 46630 }, { "grad_norm": 0.17699775099754333, "learning_rate": 5.9630425056333186e-05, "loss": 0.0221, "step": 46640 }, { "grad_norm": 0.24690614640712738, "learning_rate": 5.961419943280422e-05, "loss": 0.0275, "step": 46650 }, { "grad_norm": 0.18460527062416077, "learning_rate": 5.959797275788087e-05, "loss": 0.0213, "step": 46660 }, { "grad_norm": 0.18226993083953857, "learning_rate": 5.958174503333765e-05, "loss": 0.0133, "step": 46670 }, { "grad_norm": 0.14687354862689972, "learning_rate": 5.9565516260949195e-05, "loss": 0.0195, "step": 46680 }, { "grad_norm": 0.281515896320343, "learning_rate": 5.9549286442490273e-05, "loss": 0.0197, "step": 46690 }, { "grad_norm": 0.13910642266273499, "learning_rate": 5.953305557973572e-05, "loss": 0.0157, "step": 46700 }, { "grad_norm": 0.29769453406333923, "learning_rate": 5.9516823674460564e-05, "loss": 0.0191, "step": 46710 }, { "grad_norm": 0.16870799660682678, "learning_rate": 5.9500590728439875e-05, "loss": 0.017, "step": 46720 }, { "grad_norm": 0.2839260995388031, "learning_rate": 5.948435674344886e-05, "loss": 0.0184, "step": 46730 }, { "grad_norm": 0.2576328217983246, "learning_rate": 5.946812172126285e-05, "loss": 0.0123, "step": 46740 }, { "grad_norm": 0.08084467053413391, "learning_rate": 5.945188566365729e-05, "loss": 0.0174, "step": 46750 }, { "grad_norm": 0.22635936737060547, "learning_rate": 5.943564857240773e-05, "loss": 0.0208, "step": 46760 }, { "grad_norm": 0.15660369396209717, "learning_rate": 5.941941044928981e-05, "loss": 0.0227, "step": 46770 }, { "grad_norm": 0.2788432240486145, "learning_rate": 5.940317129607935e-05, "loss": 0.0186, "step": 46780 }, { "grad_norm": 0.5788813233375549, "learning_rate": 5.9386931114552204e-05, "loss": 0.0137, "step": 46790 }, { "grad_norm": 0.2232261747121811, "learning_rate": 5.937068990648438e-05, "loss": 0.0178, "step": 46800 }, { "grad_norm": 0.25316399335861206, "learning_rate": 5.935444767365199e-05, "loss": 0.031, "step": 46810 }, { "grad_norm": 0.24814321100711823, "learning_rate": 5.933820441783129e-05, "loss": 0.0257, "step": 46820 }, { "grad_norm": 0.17312169075012207, "learning_rate": 5.932196014079857e-05, "loss": 0.0242, "step": 46830 }, { "grad_norm": 0.1443440318107605, "learning_rate": 5.930571484433032e-05, "loss": 0.0116, "step": 46840 }, { "grad_norm": 0.23102526366710663, "learning_rate": 5.928946853020309e-05, "loss": 0.0246, "step": 46850 }, { "grad_norm": 0.1340872198343277, "learning_rate": 5.927322120019352e-05, "loss": 0.0194, "step": 46860 }, { "grad_norm": 0.3420329988002777, "learning_rate": 5.925697285607844e-05, "loss": 0.0198, "step": 46870 }, { "grad_norm": 0.22726504504680634, "learning_rate": 5.924072349963472e-05, "loss": 0.0188, "step": 46880 }, { "grad_norm": 0.16333870589733124, "learning_rate": 5.922447313263938e-05, "loss": 0.0145, "step": 46890 }, { "grad_norm": 0.3916093707084656, "learning_rate": 5.920822175686952e-05, "loss": 0.0199, "step": 46900 }, { "grad_norm": 0.2832273244857788, "learning_rate": 5.919196937410239e-05, "loss": 0.0206, "step": 46910 }, { "grad_norm": 0.18208225071430206, "learning_rate": 5.917571598611529e-05, "loss": 0.0203, "step": 46920 }, { "grad_norm": 0.25692129135131836, "learning_rate": 5.9159461594685686e-05, "loss": 0.0175, "step": 46930 }, { "grad_norm": 0.32216593623161316, "learning_rate": 5.914320620159114e-05, "loss": 0.0252, "step": 46940 }, { "grad_norm": 0.46636199951171875, "learning_rate": 5.912694980860931e-05, "loss": 0.0181, "step": 46950 }, { "grad_norm": 0.45756450295448303, "learning_rate": 5.9110692417517964e-05, "loss": 0.0215, "step": 46960 }, { "grad_norm": 0.3290334939956665, "learning_rate": 5.9094434030095e-05, "loss": 0.0195, "step": 46970 }, { "grad_norm": 0.19195228815078735, "learning_rate": 5.907817464811841e-05, "loss": 0.0165, "step": 46980 }, { "grad_norm": 0.3482156991958618, "learning_rate": 5.906191427336627e-05, "loss": 0.0142, "step": 46990 }, { "grad_norm": 0.29138171672821045, "learning_rate": 5.9045652907616835e-05, "loss": 0.0203, "step": 47000 }, { "grad_norm": 0.7720693945884705, "learning_rate": 5.902939055264838e-05, "loss": 0.0199, "step": 47010 }, { "grad_norm": 0.1472695916891098, "learning_rate": 5.901312721023935e-05, "loss": 0.0196, "step": 47020 }, { "grad_norm": 0.2403741478919983, "learning_rate": 5.8996862882168294e-05, "loss": 0.0176, "step": 47030 }, { "grad_norm": 0.2538537085056305, "learning_rate": 5.8980597570213826e-05, "loss": 0.021, "step": 47040 }, { "grad_norm": 0.14702658355236053, "learning_rate": 5.896433127615471e-05, "loss": 0.017, "step": 47050 }, { "grad_norm": 0.3855397403240204, "learning_rate": 5.894806400176981e-05, "loss": 0.0159, "step": 47060 }, { "grad_norm": 0.1338299661874771, "learning_rate": 5.893179574883808e-05, "loss": 0.0185, "step": 47070 }, { "grad_norm": 0.18234682083129883, "learning_rate": 5.8915526519138585e-05, "loss": 0.0119, "step": 47080 }, { "grad_norm": 0.7458483576774597, "learning_rate": 5.889925631445053e-05, "loss": 0.0228, "step": 47090 }, { "grad_norm": 0.09491806477308273, "learning_rate": 5.888298513655318e-05, "loss": 0.0181, "step": 47100 }, { "grad_norm": 0.2299140840768814, "learning_rate": 5.886671298722592e-05, "loss": 0.0209, "step": 47110 }, { "grad_norm": 0.12041782587766647, "learning_rate": 5.885043986824828e-05, "loss": 0.0131, "step": 47120 }, { "grad_norm": 0.2527570426464081, "learning_rate": 5.883416578139982e-05, "loss": 0.0161, "step": 47130 }, { "grad_norm": 0.35028815269470215, "learning_rate": 5.881789072846029e-05, "loss": 0.0255, "step": 47140 }, { "grad_norm": 0.4044608473777771, "learning_rate": 5.8801614711209474e-05, "loss": 0.0203, "step": 47150 }, { "grad_norm": 0.2731383740901947, "learning_rate": 5.878533773142732e-05, "loss": 0.0194, "step": 47160 }, { "grad_norm": 0.35514047741889954, "learning_rate": 5.8769059790893834e-05, "loss": 0.02, "step": 47170 }, { "grad_norm": 0.19932523369789124, "learning_rate": 5.8752780891389145e-05, "loss": 0.0264, "step": 47180 }, { "grad_norm": 0.3069203794002533, "learning_rate": 5.873650103469351e-05, "loss": 0.02, "step": 47190 }, { "grad_norm": 0.09885910898447037, "learning_rate": 5.872022022258725e-05, "loss": 0.0123, "step": 47200 }, { "grad_norm": 0.21190395951271057, "learning_rate": 5.8703938456850814e-05, "loss": 0.0163, "step": 47210 }, { "grad_norm": 0.37989941239356995, "learning_rate": 5.868765573926477e-05, "loss": 0.0185, "step": 47220 }, { "grad_norm": 0.3981318175792694, "learning_rate": 5.867137207160974e-05, "loss": 0.0177, "step": 47230 }, { "grad_norm": 0.15946535766124725, "learning_rate": 5.86550874556665e-05, "loss": 0.0229, "step": 47240 }, { "grad_norm": 0.35065698623657227, "learning_rate": 5.863880189321592e-05, "loss": 0.0196, "step": 47250 }, { "grad_norm": 0.3415113687515259, "learning_rate": 5.862251538603894e-05, "loss": 0.0316, "step": 47260 }, { "grad_norm": 0.2619541585445404, "learning_rate": 5.8606227935916656e-05, "loss": 0.0202, "step": 47270 }, { "grad_norm": 0.26275354623794556, "learning_rate": 5.858993954463021e-05, "loss": 0.0191, "step": 47280 }, { "grad_norm": 0.2734672725200653, "learning_rate": 5.8573650213960896e-05, "loss": 0.0218, "step": 47290 }, { "grad_norm": 0.34396040439605713, "learning_rate": 5.8557359945690084e-05, "loss": 0.0187, "step": 47300 }, { "grad_norm": 0.22023549675941467, "learning_rate": 5.8541068741599246e-05, "loss": 0.0218, "step": 47310 }, { "grad_norm": 0.20002874732017517, "learning_rate": 5.8524776603469985e-05, "loss": 0.0196, "step": 47320 }, { "grad_norm": 0.09365639090538025, "learning_rate": 5.850848353308397e-05, "loss": 0.0204, "step": 47330 }, { "grad_norm": 0.19531232118606567, "learning_rate": 5.849218953222297e-05, "loss": 0.0165, "step": 47340 }, { "grad_norm": 0.18681801855564117, "learning_rate": 5.847589460266891e-05, "loss": 0.0211, "step": 47350 }, { "grad_norm": 0.23246370255947113, "learning_rate": 5.845959874620375e-05, "loss": 0.0188, "step": 47360 }, { "grad_norm": 0.5325900316238403, "learning_rate": 5.84433019646096e-05, "loss": 0.0317, "step": 47370 }, { "grad_norm": 0.33317992091178894, "learning_rate": 5.842700425966863e-05, "loss": 0.0251, "step": 47380 }, { "grad_norm": 0.24808910489082336, "learning_rate": 5.841070563316315e-05, "loss": 0.0241, "step": 47390 }, { "grad_norm": 0.10332329571247101, "learning_rate": 5.839440608687554e-05, "loss": 0.016, "step": 47400 }, { "grad_norm": 0.18429382145404816, "learning_rate": 5.837810562258831e-05, "loss": 0.0191, "step": 47410 }, { "grad_norm": 0.20178112387657166, "learning_rate": 5.836180424208405e-05, "loss": 0.0279, "step": 47420 }, { "grad_norm": 0.19268399477005005, "learning_rate": 5.834550194714543e-05, "loss": 0.0168, "step": 47430 }, { "grad_norm": 0.15046843886375427, "learning_rate": 5.832919873955528e-05, "loss": 0.0273, "step": 47440 }, { "grad_norm": 0.2174178659915924, "learning_rate": 5.831289462109648e-05, "loss": 0.0276, "step": 47450 }, { "grad_norm": 0.21734848618507385, "learning_rate": 5.829658959355202e-05, "loss": 0.0296, "step": 47460 }, { "grad_norm": 0.11213304102420807, "learning_rate": 5.8280283658704995e-05, "loss": 0.0219, "step": 47470 }, { "grad_norm": 0.19776955246925354, "learning_rate": 5.826397681833859e-05, "loss": 0.02, "step": 47480 }, { "grad_norm": 0.18690158426761627, "learning_rate": 5.824766907423612e-05, "loss": 0.0147, "step": 47490 }, { "grad_norm": 0.15239790081977844, "learning_rate": 5.8231360428180945e-05, "loss": 0.0246, "step": 47500 }, { "grad_norm": 0.28414347767829895, "learning_rate": 5.821505088195658e-05, "loss": 0.0171, "step": 47510 }, { "grad_norm": 0.1489231288433075, "learning_rate": 5.819874043734661e-05, "loss": 0.0146, "step": 47520 }, { "grad_norm": 0.21582269668579102, "learning_rate": 5.8182429096134695e-05, "loss": 0.0181, "step": 47530 }, { "grad_norm": 0.14824914932250977, "learning_rate": 5.816611686010465e-05, "loss": 0.016, "step": 47540 }, { "grad_norm": 0.2966735363006592, "learning_rate": 5.814980373104033e-05, "loss": 0.03, "step": 47550 }, { "grad_norm": 0.29253652691841125, "learning_rate": 5.813348971072572e-05, "loss": 0.0171, "step": 47560 }, { "grad_norm": 0.3903197944164276, "learning_rate": 5.811717480094492e-05, "loss": 0.0261, "step": 47570 }, { "grad_norm": 0.2874969244003296, "learning_rate": 5.810085900348209e-05, "loss": 0.0187, "step": 47580 }, { "grad_norm": 0.6072120666503906, "learning_rate": 5.8084542320121483e-05, "loss": 0.0196, "step": 47590 }, { "grad_norm": 0.25283172726631165, "learning_rate": 5.8068224752647497e-05, "loss": 0.0321, "step": 47600 }, { "grad_norm": 0.15203943848609924, "learning_rate": 5.805190630284457e-05, "loss": 0.0132, "step": 47610 }, { "grad_norm": 0.472393274307251, "learning_rate": 5.803558697249729e-05, "loss": 0.0204, "step": 47620 }, { "grad_norm": 0.5064218640327454, "learning_rate": 5.8019266763390276e-05, "loss": 0.0223, "step": 47630 }, { "grad_norm": 0.2955729067325592, "learning_rate": 5.800294567730833e-05, "loss": 0.0144, "step": 47640 }, { "grad_norm": 0.30866310000419617, "learning_rate": 5.798662371603628e-05, "loss": 0.0145, "step": 47650 }, { "grad_norm": 0.3249112367630005, "learning_rate": 5.797030088135904e-05, "loss": 0.0191, "step": 47660 }, { "grad_norm": 0.23661650717258453, "learning_rate": 5.7953977175061705e-05, "loss": 0.0136, "step": 47670 }, { "grad_norm": 0.4724227786064148, "learning_rate": 5.793765259892938e-05, "loss": 0.0225, "step": 47680 }, { "grad_norm": 0.1726418435573578, "learning_rate": 5.792132715474729e-05, "loss": 0.0114, "step": 47690 }, { "grad_norm": 0.16794097423553467, "learning_rate": 5.790500084430078e-05, "loss": 0.0117, "step": 47700 }, { "grad_norm": 0.11037063598632812, "learning_rate": 5.7888673669375274e-05, "loss": 0.014, "step": 47710 }, { "grad_norm": 0.362017422914505, "learning_rate": 5.787234563175625e-05, "loss": 0.0241, "step": 47720 }, { "grad_norm": 0.4184058904647827, "learning_rate": 5.7856016733229355e-05, "loss": 0.0193, "step": 47730 }, { "grad_norm": 0.2530609369277954, "learning_rate": 5.7839686975580297e-05, "loss": 0.0296, "step": 47740 }, { "grad_norm": 0.25921157002449036, "learning_rate": 5.782335636059484e-05, "loss": 0.0234, "step": 47750 }, { "grad_norm": 0.18369652330875397, "learning_rate": 5.780702489005889e-05, "loss": 0.0154, "step": 47760 }, { "grad_norm": 0.2922976613044739, "learning_rate": 5.779069256575845e-05, "loss": 0.0225, "step": 47770 }, { "grad_norm": 0.13114199042320251, "learning_rate": 5.7774359389479574e-05, "loss": 0.0199, "step": 47780 }, { "grad_norm": 0.3062310814857483, "learning_rate": 5.775802536300845e-05, "loss": 0.0193, "step": 47790 }, { "grad_norm": 0.15899614989757538, "learning_rate": 5.774169048813134e-05, "loss": 0.0182, "step": 47800 }, { "grad_norm": 0.27224040031433105, "learning_rate": 5.77253547666346e-05, "loss": 0.0127, "step": 47810 }, { "grad_norm": 0.2852644622325897, "learning_rate": 5.770901820030465e-05, "loss": 0.0165, "step": 47820 }, { "grad_norm": 0.1619144231081009, "learning_rate": 5.769268079092809e-05, "loss": 0.0148, "step": 47830 }, { "grad_norm": 0.4353320598602295, "learning_rate": 5.767634254029151e-05, "loss": 0.0236, "step": 47840 }, { "grad_norm": 0.37227535247802734, "learning_rate": 5.7660003450181655e-05, "loss": 0.0315, "step": 47850 }, { "grad_norm": 0.2979346215724945, "learning_rate": 5.764366352238534e-05, "loss": 0.0153, "step": 47860 }, { "grad_norm": 0.21342620253562927, "learning_rate": 5.7627322758689474e-05, "loss": 0.0132, "step": 47870 }, { "grad_norm": 0.10828966647386551, "learning_rate": 5.761098116088105e-05, "loss": 0.0149, "step": 47880 }, { "grad_norm": 0.11648894846439362, "learning_rate": 5.759463873074717e-05, "loss": 0.015, "step": 47890 }, { "grad_norm": 0.22441595792770386, "learning_rate": 5.757829547007504e-05, "loss": 0.0164, "step": 47900 }, { "grad_norm": 0.1694341003894806, "learning_rate": 5.756195138065189e-05, "loss": 0.0147, "step": 47910 }, { "grad_norm": 0.2738983631134033, "learning_rate": 5.754560646426511e-05, "loss": 0.0218, "step": 47920 }, { "grad_norm": 0.10993988811969757, "learning_rate": 5.752926072270216e-05, "loss": 0.0172, "step": 47930 }, { "grad_norm": 0.1718965619802475, "learning_rate": 5.7512914157750563e-05, "loss": 0.0135, "step": 47940 }, { "grad_norm": 0.1787043660879135, "learning_rate": 5.749656677119798e-05, "loss": 0.014, "step": 47950 }, { "grad_norm": 0.16815829277038574, "learning_rate": 5.7480218564832125e-05, "loss": 0.0245, "step": 47960 }, { "grad_norm": 0.13574180006980896, "learning_rate": 5.746386954044082e-05, "loss": 0.0198, "step": 47970 }, { "grad_norm": 0.3122800886631012, "learning_rate": 5.744751969981195e-05, "loss": 0.0137, "step": 47980 }, { "grad_norm": 0.16770873963832855, "learning_rate": 5.7431169044733526e-05, "loss": 0.0234, "step": 47990 }, { "grad_norm": 0.2614474892616272, "learning_rate": 5.741481757699364e-05, "loss": 0.0207, "step": 48000 }, { "grad_norm": 0.19022957980632782, "learning_rate": 5.7398465298380434e-05, "loss": 0.015, "step": 48010 }, { "grad_norm": 0.2587544023990631, "learning_rate": 5.7382112210682193e-05, "loss": 0.0329, "step": 48020 }, { "grad_norm": 0.21933801472187042, "learning_rate": 5.7365758315687266e-05, "loss": 0.0179, "step": 48030 }, { "grad_norm": 0.19638940691947937, "learning_rate": 5.734940361518407e-05, "loss": 0.0201, "step": 48040 }, { "grad_norm": 0.16539575159549713, "learning_rate": 5.733304811096116e-05, "loss": 0.0128, "step": 48050 }, { "grad_norm": 0.2959398329257965, "learning_rate": 5.731669180480713e-05, "loss": 0.0224, "step": 48060 }, { "grad_norm": 0.35349079966545105, "learning_rate": 5.730033469851067e-05, "loss": 0.0273, "step": 48070 }, { "grad_norm": 0.22962503135204315, "learning_rate": 5.7283976793860607e-05, "loss": 0.0195, "step": 48080 }, { "grad_norm": 0.19541147351264954, "learning_rate": 5.726761809264577e-05, "loss": 0.0224, "step": 48090 }, { "grad_norm": 0.4988497197628021, "learning_rate": 5.7251258596655155e-05, "loss": 0.0197, "step": 48100 }, { "grad_norm": 0.17123953998088837, "learning_rate": 5.72348983076778e-05, "loss": 0.0272, "step": 48110 }, { "grad_norm": 0.13901983201503754, "learning_rate": 5.7218537227502854e-05, "loss": 0.0156, "step": 48120 }, { "grad_norm": 0.1292240470647812, "learning_rate": 5.720217535791951e-05, "loss": 0.0135, "step": 48130 }, { "grad_norm": 0.20557165145874023, "learning_rate": 5.718581270071711e-05, "loss": 0.0145, "step": 48140 }, { "grad_norm": 0.21674476563930511, "learning_rate": 5.716944925768505e-05, "loss": 0.0182, "step": 48150 }, { "grad_norm": 0.24186274409294128, "learning_rate": 5.7153085030612786e-05, "loss": 0.0199, "step": 48160 }, { "grad_norm": 0.17915374040603638, "learning_rate": 5.713672002128989e-05, "loss": 0.0226, "step": 48170 }, { "grad_norm": 0.14022791385650635, "learning_rate": 5.712035423150602e-05, "loss": 0.0185, "step": 48180 }, { "grad_norm": 0.33937203884124756, "learning_rate": 5.710398766305094e-05, "loss": 0.0194, "step": 48190 }, { "grad_norm": 0.29430967569351196, "learning_rate": 5.708762031771442e-05, "loss": 0.0199, "step": 48200 }, { "grad_norm": 0.2587914764881134, "learning_rate": 5.707125219728642e-05, "loss": 0.0151, "step": 48210 }, { "grad_norm": 0.0849488377571106, "learning_rate": 5.7054883303556905e-05, "loss": 0.0122, "step": 48220 }, { "grad_norm": 0.14727023243904114, "learning_rate": 5.703851363831595e-05, "loss": 0.0219, "step": 48230 }, { "grad_norm": 0.39858919382095337, "learning_rate": 5.702214320335374e-05, "loss": 0.0347, "step": 48240 }, { "grad_norm": 0.17797066271305084, "learning_rate": 5.7005772000460514e-05, "loss": 0.0164, "step": 48250 }, { "grad_norm": 0.11247758567333221, "learning_rate": 5.698940003142656e-05, "loss": 0.0213, "step": 48260 }, { "grad_norm": 0.15170276165008545, "learning_rate": 5.697302729804236e-05, "loss": 0.0155, "step": 48270 }, { "grad_norm": 0.20505115389823914, "learning_rate": 5.695665380209837e-05, "loss": 0.0177, "step": 48280 }, { "grad_norm": 0.19892022013664246, "learning_rate": 5.6940279545385165e-05, "loss": 0.0223, "step": 48290 }, { "grad_norm": 0.2038746476173401, "learning_rate": 5.692390452969344e-05, "loss": 0.0226, "step": 48300 }, { "grad_norm": 0.2886205017566681, "learning_rate": 5.690752875681392e-05, "loss": 0.017, "step": 48310 }, { "grad_norm": 0.22554844617843628, "learning_rate": 5.6891152228537435e-05, "loss": 0.0161, "step": 48320 }, { "grad_norm": 0.3544677197933197, "learning_rate": 5.687477494665492e-05, "loss": 0.019, "step": 48330 }, { "grad_norm": 0.11824528872966766, "learning_rate": 5.685839691295733e-05, "loss": 0.0219, "step": 48340 }, { "grad_norm": 0.4126432240009308, "learning_rate": 5.6842018129235786e-05, "loss": 0.0272, "step": 48350 }, { "grad_norm": 0.10785292088985443, "learning_rate": 5.6825638597281404e-05, "loss": 0.0125, "step": 48360 }, { "grad_norm": 0.21100449562072754, "learning_rate": 5.680925831888546e-05, "loss": 0.0223, "step": 48370 }, { "grad_norm": 0.09365317970514297, "learning_rate": 5.6792877295839274e-05, "loss": 0.0222, "step": 48380 }, { "grad_norm": 0.18961511552333832, "learning_rate": 5.6776495529934224e-05, "loss": 0.0155, "step": 48390 }, { "grad_norm": 0.14940696954727173, "learning_rate": 5.6760113022961824e-05, "loss": 0.0172, "step": 48400 }, { "grad_norm": 0.15347863733768463, "learning_rate": 5.6743729776713617e-05, "loss": 0.0157, "step": 48410 }, { "grad_norm": 0.4884850084781647, "learning_rate": 5.672734579298126e-05, "loss": 0.0146, "step": 48420 }, { "grad_norm": 0.21710163354873657, "learning_rate": 5.671096107355649e-05, "loss": 0.0186, "step": 48430 }, { "grad_norm": 0.2202952355146408, "learning_rate": 5.66945756202311e-05, "loss": 0.0128, "step": 48440 }, { "grad_norm": 0.1739782840013504, "learning_rate": 5.667818943479699e-05, "loss": 0.0108, "step": 48450 }, { "grad_norm": 0.22322623431682587, "learning_rate": 5.666180251904612e-05, "loss": 0.029, "step": 48460 }, { "grad_norm": 0.3691788911819458, "learning_rate": 5.6645414874770555e-05, "loss": 0.0241, "step": 48470 }, { "grad_norm": 0.25516852736473083, "learning_rate": 5.66290265037624e-05, "loss": 0.0218, "step": 48480 }, { "grad_norm": 0.09647427499294281, "learning_rate": 5.661263740781386e-05, "loss": 0.0146, "step": 48490 }, { "grad_norm": 0.2135878950357437, "learning_rate": 5.6596247588717254e-05, "loss": 0.0199, "step": 48500 }, { "grad_norm": 0.1104898750782013, "learning_rate": 5.6579857048264926e-05, "loss": 0.019, "step": 48510 }, { "grad_norm": 0.43528056144714355, "learning_rate": 5.6563465788249314e-05, "loss": 0.0228, "step": 48520 }, { "grad_norm": 0.2330726981163025, "learning_rate": 5.6547073810462956e-05, "loss": 0.0251, "step": 48530 }, { "grad_norm": 0.456912100315094, "learning_rate": 5.653068111669846e-05, "loss": 0.0198, "step": 48540 }, { "grad_norm": 0.12776659429073334, "learning_rate": 5.651428770874848e-05, "loss": 0.0175, "step": 48550 }, { "grad_norm": 0.1477460414171219, "learning_rate": 5.64978935884058e-05, "loss": 0.0209, "step": 48560 }, { "grad_norm": 0.5145792961120605, "learning_rate": 5.6481498757463244e-05, "loss": 0.0254, "step": 48570 }, { "grad_norm": 0.1488705724477768, "learning_rate": 5.646510321771373e-05, "loss": 0.0207, "step": 48580 }, { "grad_norm": 0.3527977168560028, "learning_rate": 5.644870697095024e-05, "loss": 0.0144, "step": 48590 }, { "grad_norm": 0.32526057958602905, "learning_rate": 5.643231001896586e-05, "loss": 0.0182, "step": 48600 }, { "grad_norm": 0.25466981530189514, "learning_rate": 5.6415912363553726e-05, "loss": 0.0508, "step": 48610 }, { "grad_norm": 0.34151512384414673, "learning_rate": 5.639951400650706e-05, "loss": 0.0187, "step": 48620 }, { "grad_norm": 0.14155791699886322, "learning_rate": 5.6383114949619165e-05, "loss": 0.0246, "step": 48630 }, { "grad_norm": 0.22592534124851227, "learning_rate": 5.636671519468342e-05, "loss": 0.0158, "step": 48640 }, { "grad_norm": 0.13660068809986115, "learning_rate": 5.635031474349327e-05, "loss": 0.0211, "step": 48650 }, { "grad_norm": 0.201181098818779, "learning_rate": 5.6333913597842246e-05, "loss": 0.0152, "step": 48660 }, { "grad_norm": 0.501144289970398, "learning_rate": 5.6317511759523955e-05, "loss": 0.0265, "step": 48670 }, { "grad_norm": 0.23831124603748322, "learning_rate": 5.630110923033207e-05, "loss": 0.0154, "step": 48680 }, { "grad_norm": 0.08543839305639267, "learning_rate": 5.628470601206036e-05, "loss": 0.0333, "step": 48690 }, { "grad_norm": 0.22684641182422638, "learning_rate": 5.626830210650266e-05, "loss": 0.0255, "step": 48700 }, { "grad_norm": 0.09100793302059174, "learning_rate": 5.625189751545285e-05, "loss": 0.0267, "step": 48710 }, { "grad_norm": 0.4373951256275177, "learning_rate": 5.6235492240704936e-05, "loss": 0.0187, "step": 48720 }, { "grad_norm": 0.2454521358013153, "learning_rate": 5.621908628405296e-05, "loss": 0.0192, "step": 48730 }, { "grad_norm": 0.20539580285549164, "learning_rate": 5.620267964729106e-05, "loss": 0.0321, "step": 48740 }, { "grad_norm": 0.31371667981147766, "learning_rate": 5.618627233221344e-05, "loss": 0.0207, "step": 48750 }, { "grad_norm": 0.13004425168037415, "learning_rate": 5.616986434061438e-05, "loss": 0.0146, "step": 48760 }, { "grad_norm": 0.14856497943401337, "learning_rate": 5.615345567428822e-05, "loss": 0.0149, "step": 48770 }, { "grad_norm": 0.22109495103359222, "learning_rate": 5.613704633502941e-05, "loss": 0.0101, "step": 48780 }, { "grad_norm": 0.33374759554862976, "learning_rate": 5.612063632463245e-05, "loss": 0.0209, "step": 48790 }, { "grad_norm": 0.3866467773914337, "learning_rate": 5.610422564489188e-05, "loss": 0.0157, "step": 48800 }, { "grad_norm": 0.6261619925498962, "learning_rate": 5.608781429760239e-05, "loss": 0.0199, "step": 48810 }, { "grad_norm": 0.19847536087036133, "learning_rate": 5.607140228455866e-05, "loss": 0.0186, "step": 48820 }, { "grad_norm": 0.2047858089208603, "learning_rate": 5.605498960755553e-05, "loss": 0.0228, "step": 48830 }, { "grad_norm": 0.2736973464488983, "learning_rate": 5.603857626838782e-05, "loss": 0.0202, "step": 48840 }, { "grad_norm": 0.15063835680484772, "learning_rate": 5.60221622688505e-05, "loss": 0.022, "step": 48850 }, { "grad_norm": 0.11231264472007751, "learning_rate": 5.6005747610738565e-05, "loss": 0.0176, "step": 48860 }, { "grad_norm": 0.34639713168144226, "learning_rate": 5.59893322958471e-05, "loss": 0.0195, "step": 48870 }, { "grad_norm": 0.3553833067417145, "learning_rate": 5.5972916325971256e-05, "loss": 0.0235, "step": 48880 }, { "grad_norm": 0.33999133110046387, "learning_rate": 5.595649970290628e-05, "loss": 0.0194, "step": 48890 }, { "grad_norm": 0.19996121525764465, "learning_rate": 5.5940082428447426e-05, "loss": 0.0241, "step": 48900 }, { "grad_norm": 0.30009451508522034, "learning_rate": 5.592366450439012e-05, "loss": 0.0267, "step": 48910 }, { "grad_norm": 0.2961830198764801, "learning_rate": 5.590724593252975e-05, "loss": 0.021, "step": 48920 }, { "grad_norm": 0.15322010219097137, "learning_rate": 5.589082671466184e-05, "loss": 0.0126, "step": 48930 }, { "grad_norm": 0.3090549111366272, "learning_rate": 5.587440685258199e-05, "loss": 0.0224, "step": 48940 }, { "grad_norm": 0.25725334882736206, "learning_rate": 5.585798634808583e-05, "loss": 0.0241, "step": 48950 }, { "grad_norm": 0.33657002449035645, "learning_rate": 5.584156520296909e-05, "loss": 0.0176, "step": 48960 }, { "grad_norm": 0.24824436008930206, "learning_rate": 5.582514341902757e-05, "loss": 0.0164, "step": 48970 }, { "grad_norm": 0.2334442287683487, "learning_rate": 5.580872099805713e-05, "loss": 0.0165, "step": 48980 }, { "grad_norm": 0.18876194953918457, "learning_rate": 5.5792297941853674e-05, "loss": 0.0132, "step": 48990 }, { "grad_norm": 0.21092413365840912, "learning_rate": 5.5775874252213247e-05, "loss": 0.0204, "step": 49000 }, { "grad_norm": 0.28556326031684875, "learning_rate": 5.575944993093189e-05, "loss": 0.0294, "step": 49010 }, { "grad_norm": 0.22066117823123932, "learning_rate": 5.574302497980574e-05, "loss": 0.0198, "step": 49020 }, { "grad_norm": 0.339774489402771, "learning_rate": 5.5726599400631005e-05, "loss": 0.0162, "step": 49030 }, { "grad_norm": 0.21838824450969696, "learning_rate": 5.571017319520399e-05, "loss": 0.013, "step": 49040 }, { "grad_norm": 0.13876953721046448, "learning_rate": 5.5693746365320984e-05, "loss": 0.0176, "step": 49050 }, { "grad_norm": 0.13103747367858887, "learning_rate": 5.567731891277848e-05, "loss": 0.0165, "step": 49060 }, { "grad_norm": 0.08895637094974518, "learning_rate": 5.566089083937288e-05, "loss": 0.0175, "step": 49070 }, { "grad_norm": 0.4230039715766907, "learning_rate": 5.564446214690079e-05, "loss": 0.0227, "step": 49080 }, { "grad_norm": 0.24856393039226532, "learning_rate": 5.56280328371588e-05, "loss": 0.0128, "step": 49090 }, { "grad_norm": 0.09877113252878189, "learning_rate": 5.5611602911943596e-05, "loss": 0.0175, "step": 49100 }, { "grad_norm": 0.293028861284256, "learning_rate": 5.559517237305194e-05, "loss": 0.0188, "step": 49110 }, { "grad_norm": 0.32113000750541687, "learning_rate": 5.557874122228064e-05, "loss": 0.0174, "step": 49120 }, { "grad_norm": 0.20769476890563965, "learning_rate": 5.55623094614266e-05, "loss": 0.0153, "step": 49130 }, { "grad_norm": 0.40741100907325745, "learning_rate": 5.5545877092286743e-05, "loss": 0.0144, "step": 49140 }, { "grad_norm": 0.33472922444343567, "learning_rate": 5.552944411665811e-05, "loss": 0.0203, "step": 49150 }, { "grad_norm": 0.2329673320055008, "learning_rate": 5.5513010536337795e-05, "loss": 0.0195, "step": 49160 }, { "grad_norm": 0.2543943226337433, "learning_rate": 5.5496576353122934e-05, "loss": 0.0189, "step": 49170 }, { "grad_norm": 0.24367886781692505, "learning_rate": 5.548014156881074e-05, "loss": 0.0128, "step": 49180 }, { "grad_norm": 0.2262413650751114, "learning_rate": 5.5463706185198494e-05, "loss": 0.0247, "step": 49190 }, { "grad_norm": 0.21559496223926544, "learning_rate": 5.544727020408358e-05, "loss": 0.0166, "step": 49200 }, { "grad_norm": 0.15655240416526794, "learning_rate": 5.5430833627263366e-05, "loss": 0.0144, "step": 49210 }, { "grad_norm": 0.3134090304374695, "learning_rate": 5.5414396456535354e-05, "loss": 0.0204, "step": 49220 }, { "grad_norm": 0.45788848400115967, "learning_rate": 5.539795869369709e-05, "loss": 0.0246, "step": 49230 }, { "grad_norm": 0.20981644093990326, "learning_rate": 5.538152034054618e-05, "loss": 0.0163, "step": 49240 }, { "grad_norm": 0.12945961952209473, "learning_rate": 5.536508139888028e-05, "loss": 0.0163, "step": 49250 }, { "grad_norm": 0.16516752541065216, "learning_rate": 5.534864187049716e-05, "loss": 0.022, "step": 49260 }, { "grad_norm": 0.1859825700521469, "learning_rate": 5.5332201757194615e-05, "loss": 0.014, "step": 49270 }, { "grad_norm": 0.31070828437805176, "learning_rate": 5.531576106077049e-05, "loss": 0.0184, "step": 49280 }, { "grad_norm": 0.18844066560268402, "learning_rate": 5.529931978302272e-05, "loss": 0.0251, "step": 49290 }, { "grad_norm": 0.24877984821796417, "learning_rate": 5.528287792574932e-05, "loss": 0.014, "step": 49300 }, { "grad_norm": 0.3616034686565399, "learning_rate": 5.5266435490748324e-05, "loss": 0.0137, "step": 49310 }, { "grad_norm": 0.3541339635848999, "learning_rate": 5.524999247981787e-05, "loss": 0.0199, "step": 49320 }, { "grad_norm": 0.11814405024051666, "learning_rate": 5.523354889475613e-05, "loss": 0.0184, "step": 49330 }, { "grad_norm": 0.25759127736091614, "learning_rate": 5.521710473736134e-05, "loss": 0.0236, "step": 49340 }, { "grad_norm": 0.42123427987098694, "learning_rate": 5.5200660009431826e-05, "loss": 0.0203, "step": 49350 }, { "grad_norm": 0.3830575942993164, "learning_rate": 5.518421471276596e-05, "loss": 0.0157, "step": 49360 }, { "grad_norm": 0.2338254302740097, "learning_rate": 5.5167768849162174e-05, "loss": 0.0178, "step": 49370 }, { "grad_norm": 0.21909582614898682, "learning_rate": 5.515132242041893e-05, "loss": 0.021, "step": 49380 }, { "grad_norm": 0.18980072438716888, "learning_rate": 5.513487542833483e-05, "loss": 0.0189, "step": 49390 }, { "grad_norm": 0.1338760107755661, "learning_rate": 5.5118427874708466e-05, "loss": 0.0151, "step": 49400 }, { "grad_norm": 0.14438778162002563, "learning_rate": 5.510197976133853e-05, "loss": 0.0148, "step": 49410 }, { "grad_norm": 0.19170735776424408, "learning_rate": 5.508553109002376e-05, "loss": 0.0121, "step": 49420 }, { "grad_norm": 0.2736513614654541, "learning_rate": 5.5069081862562957e-05, "loss": 0.0166, "step": 49430 }, { "grad_norm": 0.3202003538608551, "learning_rate": 5.5052632080754965e-05, "loss": 0.0145, "step": 49440 }, { "grad_norm": 0.07634080946445465, "learning_rate": 5.503618174639874e-05, "loss": 0.0151, "step": 49450 }, { "grad_norm": 0.2580653727054596, "learning_rate": 5.501973086129325e-05, "loss": 0.0184, "step": 49460 }, { "grad_norm": 0.34225985407829285, "learning_rate": 5.500327942723753e-05, "loss": 0.0141, "step": 49470 }, { "grad_norm": 0.14543220400810242, "learning_rate": 5.498682744603071e-05, "loss": 0.015, "step": 49480 }, { "grad_norm": 0.30565622448921204, "learning_rate": 5.497037491947194e-05, "loss": 0.0151, "step": 49490 }, { "grad_norm": 0.24408820271492004, "learning_rate": 5.4953921849360424e-05, "loss": 0.0214, "step": 49500 }, { "grad_norm": 0.295259565114975, "learning_rate": 5.493746823749547e-05, "loss": 0.0224, "step": 49510 }, { "grad_norm": 0.2146729677915573, "learning_rate": 5.49210140856764e-05, "loss": 0.0261, "step": 49520 }, { "grad_norm": 0.2469795197248459, "learning_rate": 5.4904559395702647e-05, "loss": 0.0179, "step": 49530 }, { "grad_norm": 0.18650217354297638, "learning_rate": 5.488810416937364e-05, "loss": 0.0081, "step": 49540 }, { "grad_norm": 0.33076679706573486, "learning_rate": 5.4871648408488926e-05, "loss": 0.0245, "step": 49550 }, { "grad_norm": 0.11570163816213608, "learning_rate": 5.485519211484807e-05, "loss": 0.0106, "step": 49560 }, { "grad_norm": 0.3815346956253052, "learning_rate": 5.4838735290250676e-05, "loss": 0.0226, "step": 49570 }, { "grad_norm": 0.33885258436203003, "learning_rate": 5.482227793649649e-05, "loss": 0.0181, "step": 49580 }, { "grad_norm": 0.3070630729198456, "learning_rate": 5.480582005538524e-05, "loss": 0.0176, "step": 49590 }, { "grad_norm": 0.18608878552913666, "learning_rate": 5.478936164871671e-05, "loss": 0.021, "step": 49600 }, { "grad_norm": 0.3513111472129822, "learning_rate": 5.477290271829081e-05, "loss": 0.0215, "step": 49610 }, { "grad_norm": 0.2303006947040558, "learning_rate": 5.475644326590744e-05, "loss": 0.015, "step": 49620 }, { "grad_norm": 0.09072630107402802, "learning_rate": 5.473998329336658e-05, "loss": 0.022, "step": 49630 }, { "grad_norm": 0.454712837934494, "learning_rate": 5.4723522802468286e-05, "loss": 0.0247, "step": 49640 }, { "grad_norm": 0.17185020446777344, "learning_rate": 5.470706179501264e-05, "loss": 0.0152, "step": 49650 }, { "grad_norm": 0.13981761038303375, "learning_rate": 5.4690600272799776e-05, "loss": 0.0216, "step": 49660 }, { "grad_norm": 0.09638936072587967, "learning_rate": 5.467413823762993e-05, "loss": 0.0148, "step": 49670 }, { "grad_norm": 0.35786181688308716, "learning_rate": 5.465767569130335e-05, "loss": 0.0229, "step": 49680 }, { "grad_norm": 0.17270544171333313, "learning_rate": 5.464121263562036e-05, "loss": 0.0222, "step": 49690 }, { "grad_norm": 0.22382287681102753, "learning_rate": 5.4624749072381343e-05, "loss": 0.0217, "step": 49700 }, { "grad_norm": 0.15934044122695923, "learning_rate": 5.460828500338672e-05, "loss": 0.0168, "step": 49710 }, { "grad_norm": 0.2285441756248474, "learning_rate": 5.459182043043698e-05, "loss": 0.0295, "step": 49720 }, { "grad_norm": 0.19164785742759705, "learning_rate": 5.457535535533265e-05, "loss": 0.0152, "step": 49730 }, { "grad_norm": 0.18251129984855652, "learning_rate": 5.4558889779874334e-05, "loss": 0.0153, "step": 49740 }, { "grad_norm": 0.20177912712097168, "learning_rate": 5.454242370586269e-05, "loss": 0.0222, "step": 49750 }, { "grad_norm": 0.16729646921157837, "learning_rate": 5.452595713509843e-05, "loss": 0.0247, "step": 49760 }, { "grad_norm": 0.06953786313533783, "learning_rate": 5.4509490069382274e-05, "loss": 0.0236, "step": 49770 }, { "grad_norm": 0.20405399799346924, "learning_rate": 5.449302251051509e-05, "loss": 0.019, "step": 49780 }, { "grad_norm": 0.2504080533981323, "learning_rate": 5.44765544602977e-05, "loss": 0.0226, "step": 49790 }, { "grad_norm": 0.17247357964515686, "learning_rate": 5.446008592053107e-05, "loss": 0.0227, "step": 49800 }, { "grad_norm": 0.15722428262233734, "learning_rate": 5.4443616893016135e-05, "loss": 0.0146, "step": 49810 }, { "grad_norm": 0.11329834163188934, "learning_rate": 5.4427147379553934e-05, "loss": 0.0202, "step": 49820 }, { "grad_norm": 0.21717974543571472, "learning_rate": 5.441067738194556e-05, "loss": 0.0161, "step": 49830 }, { "grad_norm": 0.15210369229316711, "learning_rate": 5.439420690199214e-05, "loss": 0.0249, "step": 49840 }, { "grad_norm": 0.23970916867256165, "learning_rate": 5.437773594149484e-05, "loss": 0.0238, "step": 49850 }, { "grad_norm": 0.2895943820476532, "learning_rate": 5.436126450225495e-05, "loss": 0.0225, "step": 49860 }, { "grad_norm": 0.13230113685131073, "learning_rate": 5.434479258607371e-05, "loss": 0.0126, "step": 49870 }, { "grad_norm": 0.15206396579742432, "learning_rate": 5.43283201947525e-05, "loss": 0.0194, "step": 49880 }, { "grad_norm": 0.1762983351945877, "learning_rate": 5.431184733009268e-05, "loss": 0.0207, "step": 49890 }, { "grad_norm": 0.13060808181762695, "learning_rate": 5.4295373993895736e-05, "loss": 0.0144, "step": 49900 }, { "grad_norm": 0.19357000291347504, "learning_rate": 5.4278900187963157e-05, "loss": 0.0154, "step": 49910 }, { "grad_norm": 0.23519368469715118, "learning_rate": 5.426242591409646e-05, "loss": 0.0184, "step": 49920 }, { "grad_norm": 0.1420259177684784, "learning_rate": 5.42459511740973e-05, "loss": 0.014, "step": 49930 }, { "grad_norm": 0.18882973492145538, "learning_rate": 5.422947596976729e-05, "loss": 0.0172, "step": 49940 }, { "grad_norm": 0.5996400117874146, "learning_rate": 5.4213000302908134e-05, "loss": 0.0198, "step": 49950 }, { "grad_norm": 0.23083606362342834, "learning_rate": 5.419652417532162e-05, "loss": 0.0216, "step": 49960 }, { "grad_norm": 0.15095771849155426, "learning_rate": 5.4180047588809534e-05, "loss": 0.018, "step": 49970 }, { "grad_norm": 0.16029708087444305, "learning_rate": 5.4163570545173704e-05, "loss": 0.0139, "step": 49980 }, { "grad_norm": 0.2907799780368805, "learning_rate": 5.414709304621608e-05, "loss": 0.0233, "step": 49990 }, { "grad_norm": 0.2710466682910919, "learning_rate": 5.413061509373858e-05, "loss": 0.0236, "step": 50000 }, { "grad_norm": 0.12458006292581558, "learning_rate": 5.411413668954324e-05, "loss": 0.0167, "step": 50010 }, { "grad_norm": 0.2390262633562088, "learning_rate": 5.4097657835432083e-05, "loss": 0.0136, "step": 50020 }, { "grad_norm": 0.16405487060546875, "learning_rate": 5.408117853320723e-05, "loss": 0.0192, "step": 50030 }, { "grad_norm": 0.24432817101478577, "learning_rate": 5.406469878467084e-05, "loss": 0.0231, "step": 50040 }, { "grad_norm": 0.22110143303871155, "learning_rate": 5.404821859162509e-05, "loss": 0.0232, "step": 50050 }, { "grad_norm": 0.3114880919456482, "learning_rate": 5.403173795587225e-05, "loss": 0.0204, "step": 50060 }, { "grad_norm": 0.3492976725101471, "learning_rate": 5.4015256879214606e-05, "loss": 0.0144, "step": 50070 }, { "grad_norm": 0.17710146307945251, "learning_rate": 5.39987753634545e-05, "loss": 0.017, "step": 50080 }, { "grad_norm": 0.21211092174053192, "learning_rate": 5.3982293410394335e-05, "loss": 0.0142, "step": 50090 }, { "grad_norm": 0.2254522740840912, "learning_rate": 5.396581102183655e-05, "loss": 0.0188, "step": 50100 }, { "grad_norm": 0.35358327627182007, "learning_rate": 5.3949328199583634e-05, "loss": 0.0134, "step": 50110 }, { "grad_norm": 0.23276083171367645, "learning_rate": 5.393284494543813e-05, "loss": 0.0137, "step": 50120 }, { "grad_norm": 0.1926681101322174, "learning_rate": 5.391636126120262e-05, "loss": 0.03, "step": 50130 }, { "grad_norm": 0.2126033455133438, "learning_rate": 5.389987714867971e-05, "loss": 0.0155, "step": 50140 }, { "grad_norm": 0.2502221167087555, "learning_rate": 5.3883392609672114e-05, "loss": 0.0158, "step": 50150 }, { "grad_norm": 0.24462848901748657, "learning_rate": 5.3866907645982545e-05, "loss": 0.0205, "step": 50160 }, { "grad_norm": 0.24477563798427582, "learning_rate": 5.385042225941375e-05, "loss": 0.0116, "step": 50170 }, { "grad_norm": 0.25728511810302734, "learning_rate": 5.3833936451768585e-05, "loss": 0.0131, "step": 50180 }, { "grad_norm": 0.47554677724838257, "learning_rate": 5.3817450224849884e-05, "loss": 0.0173, "step": 50190 }, { "grad_norm": 0.15823613107204437, "learning_rate": 5.380096358046056e-05, "loss": 0.016, "step": 50200 }, { "grad_norm": 0.15814624726772308, "learning_rate": 5.378447652040359e-05, "loss": 0.0142, "step": 50210 }, { "grad_norm": 0.2334839105606079, "learning_rate": 5.376798904648195e-05, "loss": 0.0202, "step": 50220 }, { "grad_norm": 0.2948513329029083, "learning_rate": 5.375150116049869e-05, "loss": 0.0216, "step": 50230 }, { "grad_norm": 0.2012159824371338, "learning_rate": 5.373501286425691e-05, "loss": 0.0096, "step": 50240 }, { "grad_norm": 0.21420496702194214, "learning_rate": 5.3718524159559726e-05, "loss": 0.0085, "step": 50250 }, { "grad_norm": 0.13711604475975037, "learning_rate": 5.370203504821034e-05, "loss": 0.0262, "step": 50260 }, { "grad_norm": 0.36226439476013184, "learning_rate": 5.368554553201196e-05, "loss": 0.0231, "step": 50270 }, { "grad_norm": 0.18958412110805511, "learning_rate": 5.3669055612767874e-05, "loss": 0.0209, "step": 50280 }, { "grad_norm": 0.206179678440094, "learning_rate": 5.3652565292281376e-05, "loss": 0.0168, "step": 50290 }, { "grad_norm": 0.1835469901561737, "learning_rate": 5.3636074572355824e-05, "loss": 0.0156, "step": 50300 }, { "grad_norm": 0.14465059340000153, "learning_rate": 5.361958345479463e-05, "loss": 0.0175, "step": 50310 }, { "grad_norm": 0.12795960903167725, "learning_rate": 5.360309194140123e-05, "loss": 0.0194, "step": 50320 }, { "grad_norm": 0.20420873165130615, "learning_rate": 5.3586600033979106e-05, "loss": 0.0155, "step": 50330 }, { "grad_norm": 0.22755621373653412, "learning_rate": 5.357010773433181e-05, "loss": 0.0202, "step": 50340 }, { "grad_norm": 0.27643686532974243, "learning_rate": 5.3553615044262895e-05, "loss": 0.0163, "step": 50350 }, { "grad_norm": 0.23542773723602295, "learning_rate": 5.3537121965575964e-05, "loss": 0.0155, "step": 50360 }, { "grad_norm": 0.40171369910240173, "learning_rate": 5.352062850007471e-05, "loss": 0.016, "step": 50370 }, { "grad_norm": 0.17385804653167725, "learning_rate": 5.350413464956282e-05, "loss": 0.0123, "step": 50380 }, { "grad_norm": 0.2178592085838318, "learning_rate": 5.348764041584403e-05, "loss": 0.015, "step": 50390 }, { "grad_norm": 0.31251171231269836, "learning_rate": 5.347114580072212e-05, "loss": 0.0194, "step": 50400 }, { "grad_norm": 0.3448055684566498, "learning_rate": 5.345465080600093e-05, "loss": 0.0218, "step": 50410 }, { "grad_norm": 0.34566396474838257, "learning_rate": 5.3438155433484326e-05, "loss": 0.0224, "step": 50420 }, { "grad_norm": 0.19622448086738586, "learning_rate": 5.3421659684976197e-05, "loss": 0.0177, "step": 50430 }, { "grad_norm": 0.18700307607650757, "learning_rate": 5.340516356228052e-05, "loss": 0.0195, "step": 50440 }, { "grad_norm": 0.21194276213645935, "learning_rate": 5.338866706720128e-05, "loss": 0.0183, "step": 50450 }, { "grad_norm": 0.2677503824234009, "learning_rate": 5.337217020154249e-05, "loss": 0.0237, "step": 50460 }, { "grad_norm": 0.23939454555511475, "learning_rate": 5.335567296710825e-05, "loss": 0.017, "step": 50470 }, { "grad_norm": 0.17966187000274658, "learning_rate": 5.333917536570265e-05, "loss": 0.013, "step": 50480 }, { "grad_norm": 0.22434240579605103, "learning_rate": 5.332267739912986e-05, "loss": 0.0175, "step": 50490 }, { "grad_norm": 0.11520518362522125, "learning_rate": 5.330617906919405e-05, "loss": 0.0232, "step": 50500 }, { "grad_norm": 0.3201884329319, "learning_rate": 5.328968037769949e-05, "loss": 0.0183, "step": 50510 }, { "grad_norm": 0.19561533629894257, "learning_rate": 5.327318132645042e-05, "loss": 0.018, "step": 50520 }, { "grad_norm": 0.2124880850315094, "learning_rate": 5.325668191725116e-05, "loss": 0.0176, "step": 50530 }, { "grad_norm": 0.24316750466823578, "learning_rate": 5.324018215190606e-05, "loss": 0.0206, "step": 50540 }, { "grad_norm": 0.1193995401263237, "learning_rate": 5.3223682032219515e-05, "loss": 0.0234, "step": 50550 }, { "grad_norm": 0.1596549153327942, "learning_rate": 5.320718155999595e-05, "loss": 0.0215, "step": 50560 }, { "grad_norm": 0.1754351258277893, "learning_rate": 5.3190680737039835e-05, "loss": 0.0164, "step": 50570 }, { "grad_norm": 0.4928409457206726, "learning_rate": 5.317417956515567e-05, "loss": 0.0176, "step": 50580 }, { "grad_norm": 0.16716128587722778, "learning_rate": 5.315767804614798e-05, "loss": 0.0155, "step": 50590 }, { "grad_norm": 0.10844238102436066, "learning_rate": 5.3141176181821395e-05, "loss": 0.02, "step": 50600 }, { "grad_norm": 0.17366211116313934, "learning_rate": 5.31246739739805e-05, "loss": 0.0179, "step": 50610 }, { "grad_norm": 0.22298960387706757, "learning_rate": 5.310817142442994e-05, "loss": 0.0226, "step": 50620 }, { "grad_norm": 0.2840794324874878, "learning_rate": 5.309166853497445e-05, "loss": 0.0222, "step": 50630 }, { "grad_norm": 0.17720088362693787, "learning_rate": 5.307516530741873e-05, "loss": 0.0141, "step": 50640 }, { "grad_norm": 0.25652754306793213, "learning_rate": 5.305866174356754e-05, "loss": 0.0198, "step": 50650 }, { "grad_norm": 0.4359585642814636, "learning_rate": 5.304215784522571e-05, "loss": 0.0244, "step": 50660 }, { "grad_norm": 0.1877247393131256, "learning_rate": 5.302565361419808e-05, "loss": 0.0197, "step": 50670 }, { "grad_norm": 0.25381162762641907, "learning_rate": 5.3009149052289507e-05, "loss": 0.0279, "step": 50680 }, { "grad_norm": 0.22185947000980377, "learning_rate": 5.299264416130493e-05, "loss": 0.0151, "step": 50690 }, { "grad_norm": 0.1890973448753357, "learning_rate": 5.297613894304928e-05, "loss": 0.0128, "step": 50700 }, { "grad_norm": 0.21919649839401245, "learning_rate": 5.2959633399327534e-05, "loss": 0.0154, "step": 50710 }, { "grad_norm": 0.09444014728069305, "learning_rate": 5.294312753194476e-05, "loss": 0.018, "step": 50720 }, { "grad_norm": 0.2687767446041107, "learning_rate": 5.292662134270596e-05, "loss": 0.0195, "step": 50730 }, { "grad_norm": 0.11300882697105408, "learning_rate": 5.291011483341626e-05, "loss": 0.0246, "step": 50740 }, { "grad_norm": 0.24976800382137299, "learning_rate": 5.2893608005880767e-05, "loss": 0.0243, "step": 50750 }, { "grad_norm": 0.4672611951828003, "learning_rate": 5.287710086190467e-05, "loss": 0.0267, "step": 50760 }, { "grad_norm": 0.2684388756752014, "learning_rate": 5.286059340329314e-05, "loss": 0.0166, "step": 50770 }, { "grad_norm": 0.613961935043335, "learning_rate": 5.284408563185141e-05, "loss": 0.0289, "step": 50780 }, { "grad_norm": 0.33901384472846985, "learning_rate": 5.282757754938476e-05, "loss": 0.0133, "step": 50790 }, { "grad_norm": 0.2569085359573364, "learning_rate": 5.281106915769849e-05, "loss": 0.0269, "step": 50800 }, { "grad_norm": 0.2525906264781952, "learning_rate": 5.2794560458597897e-05, "loss": 0.0183, "step": 50810 }, { "grad_norm": 0.21585871279239655, "learning_rate": 5.27780514538884e-05, "loss": 0.0165, "step": 50820 }, { "grad_norm": 0.1164342388510704, "learning_rate": 5.2761542145375365e-05, "loss": 0.0161, "step": 50830 }, { "grad_norm": 0.08447837829589844, "learning_rate": 5.274503253486421e-05, "loss": 0.0197, "step": 50840 }, { "grad_norm": 0.29120588302612305, "learning_rate": 5.272852262416046e-05, "loss": 0.014, "step": 50850 }, { "grad_norm": 0.21080493927001953, "learning_rate": 5.2712012415069555e-05, "loss": 0.0146, "step": 50860 }, { "grad_norm": 0.26393628120422363, "learning_rate": 5.269550190939705e-05, "loss": 0.0196, "step": 50870 }, { "grad_norm": 0.3139159083366394, "learning_rate": 5.267899110894852e-05, "loss": 0.0151, "step": 50880 }, { "grad_norm": 0.12742674350738525, "learning_rate": 5.266248001552955e-05, "loss": 0.019, "step": 50890 }, { "grad_norm": 0.15029595792293549, "learning_rate": 5.264596863094575e-05, "loss": 0.024, "step": 50900 }, { "grad_norm": 0.23676849901676178, "learning_rate": 5.262945695700282e-05, "loss": 0.0163, "step": 50910 }, { "grad_norm": 0.5374495983123779, "learning_rate": 5.261294499550643e-05, "loss": 0.0183, "step": 50920 }, { "grad_norm": 0.267957866191864, "learning_rate": 5.25964327482623e-05, "loss": 0.02, "step": 50930 }, { "grad_norm": 0.3182373344898224, "learning_rate": 5.257992021707617e-05, "loss": 0.0176, "step": 50940 }, { "grad_norm": 0.2186872661113739, "learning_rate": 5.256340740375387e-05, "loss": 0.0305, "step": 50950 }, { "grad_norm": 0.18900607526302338, "learning_rate": 5.254689431010117e-05, "loss": 0.0167, "step": 50960 }, { "grad_norm": 0.14076638221740723, "learning_rate": 5.253038093792395e-05, "loss": 0.0129, "step": 50970 }, { "grad_norm": 0.1466633677482605, "learning_rate": 5.251386728902806e-05, "loss": 0.0146, "step": 50980 }, { "grad_norm": 0.1347944438457489, "learning_rate": 5.2497353365219446e-05, "loss": 0.0196, "step": 50990 }, { "grad_norm": 0.24719107151031494, "learning_rate": 5.2480839168304e-05, "loss": 0.0187, "step": 51000 }, { "grad_norm": 0.13121120631694794, "learning_rate": 5.246432470008773e-05, "loss": 0.0189, "step": 51010 }, { "grad_norm": 0.12840688228607178, "learning_rate": 5.2447809962376616e-05, "loss": 0.0189, "step": 51020 }, { "grad_norm": 0.1522091031074524, "learning_rate": 5.243129495697666e-05, "loss": 0.0142, "step": 51030 }, { "grad_norm": 0.16114367544651031, "learning_rate": 5.241477968569397e-05, "loss": 0.0207, "step": 51040 }, { "grad_norm": 0.4917023777961731, "learning_rate": 5.23982641503346e-05, "loss": 0.0228, "step": 51050 }, { "grad_norm": 0.2788936495780945, "learning_rate": 5.238174835270465e-05, "loss": 0.0207, "step": 51060 }, { "grad_norm": 0.18952009081840515, "learning_rate": 5.23652322946103e-05, "loss": 0.0146, "step": 51070 }, { "grad_norm": 0.11337724328041077, "learning_rate": 5.234871597785771e-05, "loss": 0.0152, "step": 51080 }, { "grad_norm": 0.3054770827293396, "learning_rate": 5.233219940425306e-05, "loss": 0.0186, "step": 51090 }, { "grad_norm": 0.30018359422683716, "learning_rate": 5.2315682575602576e-05, "loss": 0.0302, "step": 51100 }, { "grad_norm": 0.3575909435749054, "learning_rate": 5.229916549371254e-05, "loss": 0.0181, "step": 51110 }, { "grad_norm": 0.33862343430519104, "learning_rate": 5.228264816038924e-05, "loss": 0.0168, "step": 51120 }, { "grad_norm": 0.21598394215106964, "learning_rate": 5.2266130577438935e-05, "loss": 0.0201, "step": 51130 }, { "grad_norm": 0.16304826736450195, "learning_rate": 5.224961274666802e-05, "loss": 0.0212, "step": 51140 }, { "grad_norm": 0.13199637830257416, "learning_rate": 5.223309466988283e-05, "loss": 0.0224, "step": 51150 }, { "grad_norm": 0.28868064284324646, "learning_rate": 5.221657634888976e-05, "loss": 0.0219, "step": 51160 }, { "grad_norm": 0.249075248837471, "learning_rate": 5.2200057785495235e-05, "loss": 0.0292, "step": 51170 }, { "grad_norm": 0.19033974409103394, "learning_rate": 5.2183538981505706e-05, "loss": 0.0217, "step": 51180 }, { "grad_norm": 0.34758028388023376, "learning_rate": 5.216701993872762e-05, "loss": 0.0244, "step": 51190 }, { "grad_norm": 0.45519641041755676, "learning_rate": 5.215050065896752e-05, "loss": 0.0259, "step": 51200 }, { "grad_norm": 0.20242996513843536, "learning_rate": 5.2133981144031875e-05, "loss": 0.0269, "step": 51210 }, { "grad_norm": 0.2773697078227997, "learning_rate": 5.211746139572727e-05, "loss": 0.0181, "step": 51220 }, { "grad_norm": 0.22781957685947418, "learning_rate": 5.210094141586026e-05, "loss": 0.023, "step": 51230 }, { "grad_norm": 0.29834094643592834, "learning_rate": 5.2084421206237476e-05, "loss": 0.0193, "step": 51240 }, { "grad_norm": 0.2614303231239319, "learning_rate": 5.20679007686655e-05, "loss": 0.0158, "step": 51250 }, { "grad_norm": 0.11611299961805344, "learning_rate": 5.205138010495102e-05, "loss": 0.0163, "step": 51260 }, { "grad_norm": 0.2590375542640686, "learning_rate": 5.20348592169007e-05, "loss": 0.0217, "step": 51270 }, { "grad_norm": 0.34317871928215027, "learning_rate": 5.201833810632123e-05, "loss": 0.0238, "step": 51280 }, { "grad_norm": 0.21717330813407898, "learning_rate": 5.200181677501933e-05, "loss": 0.0161, "step": 51290 }, { "grad_norm": 0.22989898920059204, "learning_rate": 5.1985295224801777e-05, "loss": 0.0157, "step": 51300 }, { "grad_norm": 0.3335466980934143, "learning_rate": 5.196877345747531e-05, "loss": 0.0219, "step": 51310 }, { "grad_norm": 0.11555327475070953, "learning_rate": 5.195225147484674e-05, "loss": 0.0136, "step": 51320 }, { "grad_norm": 0.13567093014717102, "learning_rate": 5.193572927872288e-05, "loss": 0.0226, "step": 51330 }, { "grad_norm": 0.1500503420829773, "learning_rate": 5.191920687091059e-05, "loss": 0.0178, "step": 51340 }, { "grad_norm": 0.3707534968852997, "learning_rate": 5.1902684253216695e-05, "loss": 0.0156, "step": 51350 }, { "grad_norm": 0.33908161520957947, "learning_rate": 5.1886161427448134e-05, "loss": 0.0221, "step": 51360 }, { "grad_norm": 0.21717092394828796, "learning_rate": 5.186963839541179e-05, "loss": 0.0126, "step": 51370 }, { "grad_norm": 0.14934711158275604, "learning_rate": 5.185311515891459e-05, "loss": 0.0108, "step": 51380 }, { "grad_norm": 0.2189968377351761, "learning_rate": 5.1836591719763516e-05, "loss": 0.0157, "step": 51390 }, { "grad_norm": 0.29863500595092773, "learning_rate": 5.182006807976554e-05, "loss": 0.0184, "step": 51400 }, { "grad_norm": 0.23373475670814514, "learning_rate": 5.180354424072763e-05, "loss": 0.0213, "step": 51410 }, { "grad_norm": 0.20217058062553406, "learning_rate": 5.178702020445686e-05, "loss": 0.0134, "step": 51420 }, { "grad_norm": 0.26419970393180847, "learning_rate": 5.177049597276024e-05, "loss": 0.0137, "step": 51430 }, { "grad_norm": 0.2059413343667984, "learning_rate": 5.175397154744482e-05, "loss": 0.0183, "step": 51440 }, { "grad_norm": 0.09421586990356445, "learning_rate": 5.1737446930317736e-05, "loss": 0.0126, "step": 51450 }, { "grad_norm": 0.3221811354160309, "learning_rate": 5.172092212318604e-05, "loss": 0.0175, "step": 51460 }, { "grad_norm": 0.2614019811153412, "learning_rate": 5.170439712785692e-05, "loss": 0.017, "step": 51470 }, { "grad_norm": 0.18176265060901642, "learning_rate": 5.1687871946137467e-05, "loss": 0.0223, "step": 51480 }, { "grad_norm": 0.390271931886673, "learning_rate": 5.16713465798349e-05, "loss": 0.0188, "step": 51490 }, { "grad_norm": 0.1502838432788849, "learning_rate": 5.165482103075637e-05, "loss": 0.0152, "step": 51500 }, { "grad_norm": 0.25443533062934875, "learning_rate": 5.163829530070909e-05, "loss": 0.0181, "step": 51510 }, { "grad_norm": 0.25206494331359863, "learning_rate": 5.162176939150032e-05, "loss": 0.0237, "step": 51520 }, { "grad_norm": 0.11886338889598846, "learning_rate": 5.160524330493728e-05, "loss": 0.021, "step": 51530 }, { "grad_norm": 0.4054262936115265, "learning_rate": 5.158871704282724e-05, "loss": 0.0129, "step": 51540 }, { "grad_norm": 0.14638788998126984, "learning_rate": 5.15721906069775e-05, "loss": 0.0257, "step": 51550 }, { "grad_norm": 0.18784672021865845, "learning_rate": 5.155566399919536e-05, "loss": 0.0198, "step": 51560 }, { "grad_norm": 0.492454469203949, "learning_rate": 5.153913722128813e-05, "loss": 0.0186, "step": 51570 }, { "grad_norm": 0.25898581743240356, "learning_rate": 5.1522610275063176e-05, "loss": 0.0203, "step": 51580 }, { "grad_norm": 0.18026216328144073, "learning_rate": 5.150608316232786e-05, "loss": 0.0191, "step": 51590 }, { "grad_norm": 0.23379282653331757, "learning_rate": 5.148955588488955e-05, "loss": 0.0244, "step": 51600 }, { "grad_norm": 0.2535349428653717, "learning_rate": 5.147302844455564e-05, "loss": 0.0117, "step": 51610 }, { "grad_norm": 0.11298386752605438, "learning_rate": 5.145650084313356e-05, "loss": 0.016, "step": 51620 }, { "grad_norm": 0.11844498664140701, "learning_rate": 5.1439973082430734e-05, "loss": 0.0177, "step": 51630 }, { "grad_norm": 0.12883949279785156, "learning_rate": 5.142344516425461e-05, "loss": 0.0182, "step": 51640 }, { "grad_norm": 0.22804982960224152, "learning_rate": 5.1406917090412673e-05, "loss": 0.0227, "step": 51650 }, { "grad_norm": 0.17611359059810638, "learning_rate": 5.1390388862712404e-05, "loss": 0.0206, "step": 51660 }, { "grad_norm": 0.13247154653072357, "learning_rate": 5.137386048296129e-05, "loss": 0.0138, "step": 51670 }, { "grad_norm": 0.4340614378452301, "learning_rate": 5.135733195296686e-05, "loss": 0.0213, "step": 51680 }, { "grad_norm": 0.24106182157993317, "learning_rate": 5.134080327453664e-05, "loss": 0.0132, "step": 51690 }, { "grad_norm": 0.24128712713718414, "learning_rate": 5.1324274449478204e-05, "loss": 0.0143, "step": 51700 }, { "grad_norm": 0.29977405071258545, "learning_rate": 5.130774547959909e-05, "loss": 0.0175, "step": 51710 }, { "grad_norm": 0.4491986036300659, "learning_rate": 5.129121636670692e-05, "loss": 0.0217, "step": 51720 }, { "grad_norm": 0.27611666917800903, "learning_rate": 5.1274687112609246e-05, "loss": 0.0169, "step": 51730 }, { "grad_norm": 0.34510213136672974, "learning_rate": 5.125815771911372e-05, "loss": 0.0159, "step": 51740 }, { "grad_norm": 0.29876986145973206, "learning_rate": 5.1241628188027956e-05, "loss": 0.018, "step": 51750 }, { "grad_norm": 0.18335217237472534, "learning_rate": 5.1225098521159585e-05, "loss": 0.0157, "step": 51760 }, { "grad_norm": 0.18494349718093872, "learning_rate": 5.12085687203163e-05, "loss": 0.0198, "step": 51770 }, { "grad_norm": 0.28011465072631836, "learning_rate": 5.119203878730575e-05, "loss": 0.0189, "step": 51780 }, { "grad_norm": 0.14413943886756897, "learning_rate": 5.117550872393564e-05, "loss": 0.0219, "step": 51790 }, { "grad_norm": 0.25681325793266296, "learning_rate": 5.1158978532013645e-05, "loss": 0.0231, "step": 51800 }, { "grad_norm": 0.20706409215927124, "learning_rate": 5.114244821334751e-05, "loss": 0.0174, "step": 51810 }, { "grad_norm": 0.4341718852519989, "learning_rate": 5.112591776974497e-05, "loss": 0.0142, "step": 51820 }, { "grad_norm": 0.4611031115055084, "learning_rate": 5.110938720301374e-05, "loss": 0.0216, "step": 51830 }, { "grad_norm": 0.3428654670715332, "learning_rate": 5.1092856514961595e-05, "loss": 0.0238, "step": 51840 }, { "grad_norm": 0.1735154390335083, "learning_rate": 5.107632570739631e-05, "loss": 0.012, "step": 51850 }, { "grad_norm": 0.2536694407463074, "learning_rate": 5.1059794782125645e-05, "loss": 0.0166, "step": 51860 }, { "grad_norm": 0.2094760686159134, "learning_rate": 5.104326374095744e-05, "loss": 0.0202, "step": 51870 }, { "grad_norm": 0.2176447957754135, "learning_rate": 5.102673258569947e-05, "loss": 0.0175, "step": 51880 }, { "grad_norm": 0.20400235056877136, "learning_rate": 5.101020131815956e-05, "loss": 0.0236, "step": 51890 }, { "grad_norm": 0.3263952434062958, "learning_rate": 5.099366994014556e-05, "loss": 0.019, "step": 51900 }, { "grad_norm": 0.24900908768177032, "learning_rate": 5.097713845346532e-05, "loss": 0.0215, "step": 51910 }, { "grad_norm": 0.25470155477523804, "learning_rate": 5.096060685992668e-05, "loss": 0.0182, "step": 51920 }, { "grad_norm": 0.15302082896232605, "learning_rate": 5.0944075161337514e-05, "loss": 0.0187, "step": 51930 }, { "grad_norm": 0.35187721252441406, "learning_rate": 5.09275433595057e-05, "loss": 0.0144, "step": 51940 }, { "grad_norm": 0.20782804489135742, "learning_rate": 5.0911011456239157e-05, "loss": 0.0242, "step": 51950 }, { "grad_norm": 0.3871626853942871, "learning_rate": 5.089447945334574e-05, "loss": 0.0246, "step": 51960 }, { "grad_norm": 0.14134812355041504, "learning_rate": 5.087794735263343e-05, "loss": 0.0173, "step": 51970 }, { "grad_norm": 0.31970635056495667, "learning_rate": 5.086141515591011e-05, "loss": 0.019, "step": 51980 }, { "grad_norm": 0.30374225974082947, "learning_rate": 5.084488286498371e-05, "loss": 0.0211, "step": 51990 }, { "grad_norm": 0.22628693282604218, "learning_rate": 5.08283504816622e-05, "loss": 0.0333, "step": 52000 }, { "grad_norm": 0.09823494404554367, "learning_rate": 5.081181800775353e-05, "loss": 0.0212, "step": 52010 }, { "grad_norm": 0.26272034645080566, "learning_rate": 5.079528544506565e-05, "loss": 0.0146, "step": 52020 }, { "grad_norm": 0.19340816140174866, "learning_rate": 5.077875279540657e-05, "loss": 0.0226, "step": 52030 }, { "grad_norm": 0.39331039786338806, "learning_rate": 5.076222006058424e-05, "loss": 0.0246, "step": 52040 }, { "grad_norm": 0.37203723192214966, "learning_rate": 5.074568724240667e-05, "loss": 0.0186, "step": 52050 }, { "grad_norm": 0.2983359396457672, "learning_rate": 5.072915434268187e-05, "loss": 0.0242, "step": 52060 }, { "grad_norm": 0.16843104362487793, "learning_rate": 5.071262136321785e-05, "loss": 0.0187, "step": 52070 }, { "grad_norm": 0.3382759392261505, "learning_rate": 5.069608830582262e-05, "loss": 0.0184, "step": 52080 }, { "grad_norm": 0.28284719586372375, "learning_rate": 5.067955517230423e-05, "loss": 0.0124, "step": 52090 }, { "grad_norm": 0.21718065440654755, "learning_rate": 5.06630219644707e-05, "loss": 0.0244, "step": 52100 }, { "grad_norm": 0.20714673399925232, "learning_rate": 5.064648868413008e-05, "loss": 0.0173, "step": 52110 }, { "grad_norm": 0.16750429570674896, "learning_rate": 5.062995533309044e-05, "loss": 0.0207, "step": 52120 }, { "grad_norm": 0.135175421833992, "learning_rate": 5.061342191315983e-05, "loss": 0.021, "step": 52130 }, { "grad_norm": 0.14092673361301422, "learning_rate": 5.0596888426146325e-05, "loss": 0.0191, "step": 52140 }, { "grad_norm": 0.16784323751926422, "learning_rate": 5.058035487385798e-05, "loss": 0.016, "step": 52150 }, { "grad_norm": 0.19337323307991028, "learning_rate": 5.0563821258102915e-05, "loss": 0.0185, "step": 52160 }, { "grad_norm": 0.38594675064086914, "learning_rate": 5.05472875806892e-05, "loss": 0.0157, "step": 52170 }, { "grad_norm": 0.1867714673280716, "learning_rate": 5.053075384342494e-05, "loss": 0.0214, "step": 52180 }, { "grad_norm": 0.1821928471326828, "learning_rate": 5.0514220048118234e-05, "loss": 0.0144, "step": 52190 }, { "grad_norm": 0.14224833250045776, "learning_rate": 5.049768619657721e-05, "loss": 0.0211, "step": 52200 }, { "grad_norm": 0.11777336895465851, "learning_rate": 5.048115229060996e-05, "loss": 0.0184, "step": 52210 }, { "grad_norm": 0.2438729703426361, "learning_rate": 5.046461833202462e-05, "loss": 0.022, "step": 52220 }, { "grad_norm": 0.1463223546743393, "learning_rate": 5.044808432262932e-05, "loss": 0.0191, "step": 52230 }, { "grad_norm": 0.2972821593284607, "learning_rate": 5.043155026423219e-05, "loss": 0.0132, "step": 52240 }, { "grad_norm": 0.12534543871879578, "learning_rate": 5.041501615864137e-05, "loss": 0.0156, "step": 52250 }, { "grad_norm": 0.19816270470619202, "learning_rate": 5.039848200766502e-05, "loss": 0.018, "step": 52260 }, { "grad_norm": 0.18061412870883942, "learning_rate": 5.038194781311125e-05, "loss": 0.0154, "step": 52270 }, { "grad_norm": 0.3312093913555145, "learning_rate": 5.0365413576788264e-05, "loss": 0.0162, "step": 52280 }, { "grad_norm": 0.24777595698833466, "learning_rate": 5.0348879300504194e-05, "loss": 0.0172, "step": 52290 }, { "grad_norm": 0.15255491435527802, "learning_rate": 5.033234498606721e-05, "loss": 0.0245, "step": 52300 }, { "grad_norm": 0.3571338951587677, "learning_rate": 5.0315810635285476e-05, "loss": 0.0251, "step": 52310 }, { "grad_norm": 0.16531480848789215, "learning_rate": 5.0299276249967164e-05, "loss": 0.0171, "step": 52320 }, { "grad_norm": 0.3359980583190918, "learning_rate": 5.0282741831920454e-05, "loss": 0.0292, "step": 52330 }, { "grad_norm": 0.3176346719264984, "learning_rate": 5.026620738295351e-05, "loss": 0.0228, "step": 52340 }, { "grad_norm": 0.31267207860946655, "learning_rate": 5.024967290487455e-05, "loss": 0.0179, "step": 52350 }, { "grad_norm": 0.2784634828567505, "learning_rate": 5.023313839949172e-05, "loss": 0.022, "step": 52360 }, { "grad_norm": 0.1902206391096115, "learning_rate": 5.0216603868613224e-05, "loss": 0.0173, "step": 52370 }, { "grad_norm": 0.3770678639411926, "learning_rate": 5.020006931404726e-05, "loss": 0.0162, "step": 52380 }, { "grad_norm": 0.11362451314926147, "learning_rate": 5.0183534737602014e-05, "loss": 0.0187, "step": 52390 }, { "grad_norm": 0.2769055664539337, "learning_rate": 5.0167000141085686e-05, "loss": 0.0246, "step": 52400 }, { "grad_norm": 0.39224037528038025, "learning_rate": 5.015046552630647e-05, "loss": 0.0184, "step": 52410 }, { "grad_norm": 0.2640659213066101, "learning_rate": 5.013393089507258e-05, "loss": 0.0196, "step": 52420 }, { "grad_norm": 0.23836617171764374, "learning_rate": 5.011739624919219e-05, "loss": 0.0204, "step": 52430 }, { "grad_norm": 0.06709542125463486, "learning_rate": 5.0100861590473545e-05, "loss": 0.0102, "step": 52440 }, { "grad_norm": 0.29229745268821716, "learning_rate": 5.008432692072482e-05, "loss": 0.0193, "step": 52450 }, { "grad_norm": 0.11606456339359283, "learning_rate": 5.006779224175423e-05, "loss": 0.0259, "step": 52460 }, { "grad_norm": 0.2435421347618103, "learning_rate": 5.005125755536998e-05, "loss": 0.0166, "step": 52470 }, { "grad_norm": 0.1838793307542801, "learning_rate": 5.003472286338029e-05, "loss": 0.0181, "step": 52480 }, { "grad_norm": 0.1330803632736206, "learning_rate": 5.001818816759335e-05, "loss": 0.0407, "step": 52490 }, { "grad_norm": 0.1464986354112625, "learning_rate": 5.000165346981738e-05, "loss": 0.0143, "step": 52500 }, { "grad_norm": 0.18648932874202728, "learning_rate": 4.9985118771860595e-05, "loss": 0.0118, "step": 52510 }, { "grad_norm": 0.24876980483531952, "learning_rate": 4.996858407553119e-05, "loss": 0.0167, "step": 52520 }, { "grad_norm": 0.43780457973480225, "learning_rate": 4.99520493826374e-05, "loss": 0.0196, "step": 52530 }, { "grad_norm": 0.28423935174942017, "learning_rate": 4.9935514694987384e-05, "loss": 0.0154, "step": 52540 }, { "grad_norm": 0.2102273851633072, "learning_rate": 4.991898001438941e-05, "loss": 0.0132, "step": 52550 }, { "grad_norm": 0.28150874376296997, "learning_rate": 4.990244534265166e-05, "loss": 0.0234, "step": 52560 }, { "grad_norm": 0.1570393443107605, "learning_rate": 4.988591068158234e-05, "loss": 0.0148, "step": 52570 }, { "grad_norm": 0.21560515463352203, "learning_rate": 4.986937603298965e-05, "loss": 0.0266, "step": 52580 }, { "grad_norm": 0.14681120216846466, "learning_rate": 4.985284139868181e-05, "loss": 0.0115, "step": 52590 }, { "grad_norm": 0.2935919761657715, "learning_rate": 4.9836306780467e-05, "loss": 0.0142, "step": 52600 }, { "grad_norm": 0.22228087484836578, "learning_rate": 4.981977218015343e-05, "loss": 0.0112, "step": 52610 }, { "grad_norm": 0.11959938704967499, "learning_rate": 4.980323759954932e-05, "loss": 0.0101, "step": 52620 }, { "grad_norm": 0.124376080930233, "learning_rate": 4.978670304046284e-05, "loss": 0.016, "step": 52630 }, { "grad_norm": 0.2761903703212738, "learning_rate": 4.9770168504702195e-05, "loss": 0.0186, "step": 52640 }, { "grad_norm": 0.27057936787605286, "learning_rate": 4.975363399407559e-05, "loss": 0.0166, "step": 52650 }, { "grad_norm": 0.15145644545555115, "learning_rate": 4.9737099510391194e-05, "loss": 0.0189, "step": 52660 }, { "grad_norm": 0.30795586109161377, "learning_rate": 4.972056505545719e-05, "loss": 0.0167, "step": 52670 }, { "grad_norm": 0.20616820454597473, "learning_rate": 4.9704030631081775e-05, "loss": 0.0145, "step": 52680 }, { "grad_norm": 0.20313486456871033, "learning_rate": 4.968749623907313e-05, "loss": 0.0156, "step": 52690 }, { "grad_norm": 0.26231497526168823, "learning_rate": 4.967096188123943e-05, "loss": 0.0207, "step": 52700 }, { "grad_norm": 0.2642291188240051, "learning_rate": 4.965442755938884e-05, "loss": 0.0176, "step": 52710 }, { "grad_norm": 0.12418684363365173, "learning_rate": 4.963789327532954e-05, "loss": 0.0119, "step": 52720 }, { "grad_norm": 0.43777453899383545, "learning_rate": 4.962135903086968e-05, "loss": 0.013, "step": 52730 }, { "grad_norm": 0.22041743993759155, "learning_rate": 4.960482482781741e-05, "loss": 0.0125, "step": 52740 }, { "grad_norm": 0.2156020700931549, "learning_rate": 4.958829066798092e-05, "loss": 0.0202, "step": 52750 }, { "grad_norm": 0.48284921050071716, "learning_rate": 4.957175655316835e-05, "loss": 0.0227, "step": 52760 }, { "grad_norm": 0.2645295560359955, "learning_rate": 4.9555222485187834e-05, "loss": 0.0239, "step": 52770 }, { "grad_norm": 0.29189079999923706, "learning_rate": 4.953868846584752e-05, "loss": 0.0171, "step": 52780 }, { "grad_norm": 0.2333531528711319, "learning_rate": 4.9522154496955544e-05, "loss": 0.0259, "step": 52790 }, { "grad_norm": 0.3091665804386139, "learning_rate": 4.950562058032001e-05, "loss": 0.0264, "step": 52800 }, { "grad_norm": 0.1291733682155609, "learning_rate": 4.948908671774909e-05, "loss": 0.0207, "step": 52810 }, { "grad_norm": 0.19529446959495544, "learning_rate": 4.947255291105087e-05, "loss": 0.021, "step": 52820 }, { "grad_norm": 0.21068856120109558, "learning_rate": 4.945601916203348e-05, "loss": 0.0195, "step": 52830 }, { "grad_norm": 0.10631808638572693, "learning_rate": 4.9439485472505016e-05, "loss": 0.0145, "step": 52840 }, { "grad_norm": 0.09945978969335556, "learning_rate": 4.942295184427356e-05, "loss": 0.0182, "step": 52850 }, { "grad_norm": 0.15560898184776306, "learning_rate": 4.9406418279147234e-05, "loss": 0.0152, "step": 52860 }, { "grad_norm": 0.2742084562778473, "learning_rate": 4.9389884778934116e-05, "loss": 0.0168, "step": 52870 }, { "grad_norm": 0.27630922198295593, "learning_rate": 4.937335134544228e-05, "loss": 0.0206, "step": 52880 }, { "grad_norm": 0.24610009789466858, "learning_rate": 4.935681798047979e-05, "loss": 0.0275, "step": 52890 }, { "grad_norm": 0.18615075945854187, "learning_rate": 4.934028468585473e-05, "loss": 0.0235, "step": 52900 }, { "grad_norm": 0.29195475578308105, "learning_rate": 4.932375146337514e-05, "loss": 0.0229, "step": 52910 }, { "grad_norm": 0.18413269519805908, "learning_rate": 4.930721831484908e-05, "loss": 0.014, "step": 52920 }, { "grad_norm": 0.13451434671878815, "learning_rate": 4.929068524208456e-05, "loss": 0.0229, "step": 52930 }, { "grad_norm": 0.09900406748056412, "learning_rate": 4.927415224688965e-05, "loss": 0.0194, "step": 52940 }, { "grad_norm": 0.3476671278476715, "learning_rate": 4.9257619331072375e-05, "loss": 0.026, "step": 52950 }, { "grad_norm": 0.25666874647140503, "learning_rate": 4.924108649644072e-05, "loss": 0.0145, "step": 52960 }, { "grad_norm": 0.2606223523616791, "learning_rate": 4.922455374480271e-05, "loss": 0.0225, "step": 52970 }, { "grad_norm": 0.29391151666641235, "learning_rate": 4.920802107796634e-05, "loss": 0.0237, "step": 52980 }, { "grad_norm": 0.19617736339569092, "learning_rate": 4.9191488497739564e-05, "loss": 0.0169, "step": 52990 }, { "grad_norm": 0.15144208073616028, "learning_rate": 4.917495600593042e-05, "loss": 0.0155, "step": 53000 }, { "grad_norm": 0.2874632477760315, "learning_rate": 4.915842360434684e-05, "loss": 0.0203, "step": 53010 }, { "grad_norm": 0.15362535417079926, "learning_rate": 4.914189129479679e-05, "loss": 0.0222, "step": 53020 }, { "grad_norm": 0.4339229166507721, "learning_rate": 4.912535907908822e-05, "loss": 0.0205, "step": 53030 }, { "grad_norm": 0.258255273103714, "learning_rate": 4.910882695902907e-05, "loss": 0.0152, "step": 53040 }, { "grad_norm": 0.2818492352962494, "learning_rate": 4.9092294936427234e-05, "loss": 0.0186, "step": 53050 }, { "grad_norm": 0.31911900639533997, "learning_rate": 4.9075763013090685e-05, "loss": 0.0264, "step": 53060 }, { "grad_norm": 0.4138604998588562, "learning_rate": 4.9059231190827305e-05, "loss": 0.0146, "step": 53070 }, { "grad_norm": 0.12172561883926392, "learning_rate": 4.904269947144499e-05, "loss": 0.0291, "step": 53080 }, { "grad_norm": 0.366036057472229, "learning_rate": 4.902616785675161e-05, "loss": 0.0247, "step": 53090 }, { "grad_norm": 0.45619282126426697, "learning_rate": 4.900963634855505e-05, "loss": 0.0227, "step": 53100 }, { "grad_norm": 0.21395336091518402, "learning_rate": 4.8993104948663173e-05, "loss": 0.0201, "step": 53110 }, { "grad_norm": 0.36443424224853516, "learning_rate": 4.8976573658883823e-05, "loss": 0.0184, "step": 53120 }, { "grad_norm": 0.3043082356452942, "learning_rate": 4.896004248102484e-05, "loss": 0.0184, "step": 53130 }, { "grad_norm": 0.15533360838890076, "learning_rate": 4.8943511416894045e-05, "loss": 0.0172, "step": 53140 }, { "grad_norm": 0.35890084505081177, "learning_rate": 4.892698046829926e-05, "loss": 0.0265, "step": 53150 }, { "grad_norm": 0.2980867922306061, "learning_rate": 4.8910449637048275e-05, "loss": 0.019, "step": 53160 }, { "grad_norm": 0.20986497402191162, "learning_rate": 4.8893918924948884e-05, "loss": 0.0246, "step": 53170 }, { "grad_norm": 0.21737048029899597, "learning_rate": 4.8877388333808835e-05, "loss": 0.0237, "step": 53180 }, { "grad_norm": 0.26850929856300354, "learning_rate": 4.8860857865435925e-05, "loss": 0.0215, "step": 53190 }, { "grad_norm": 0.4212754964828491, "learning_rate": 4.884432752163789e-05, "loss": 0.0183, "step": 53200 }, { "grad_norm": 0.2895158529281616, "learning_rate": 4.8827797304222466e-05, "loss": 0.019, "step": 53210 }, { "grad_norm": 0.09964972734451294, "learning_rate": 4.8811267214997366e-05, "loss": 0.0143, "step": 53220 }, { "grad_norm": 0.1464163362979889, "learning_rate": 4.879473725577029e-05, "loss": 0.0151, "step": 53230 }, { "grad_norm": 0.19597700238227844, "learning_rate": 4.877820742834891e-05, "loss": 0.0151, "step": 53240 }, { "grad_norm": 0.11614334583282471, "learning_rate": 4.8761677734540956e-05, "loss": 0.0168, "step": 53250 }, { "grad_norm": 0.14215689897537231, "learning_rate": 4.874514817615407e-05, "loss": 0.0109, "step": 53260 }, { "grad_norm": 0.17423813045024872, "learning_rate": 4.872861875499588e-05, "loss": 0.0137, "step": 53270 }, { "grad_norm": 0.15209472179412842, "learning_rate": 4.871208947287404e-05, "loss": 0.0174, "step": 53280 }, { "grad_norm": 0.2831672132015228, "learning_rate": 4.869556033159615e-05, "loss": 0.0239, "step": 53290 }, { "grad_norm": 0.29222816228866577, "learning_rate": 4.8679031332969835e-05, "loss": 0.0195, "step": 53300 }, { "grad_norm": 0.23220060765743256, "learning_rate": 4.8662502478802635e-05, "loss": 0.0236, "step": 53310 }, { "grad_norm": 0.2786187529563904, "learning_rate": 4.8645973770902176e-05, "loss": 0.0252, "step": 53320 }, { "grad_norm": 0.20993037521839142, "learning_rate": 4.862944521107599e-05, "loss": 0.0166, "step": 53330 }, { "grad_norm": 0.19859127700328827, "learning_rate": 4.861291680113162e-05, "loss": 0.0141, "step": 53340 }, { "grad_norm": 0.20533519983291626, "learning_rate": 4.8596388542876563e-05, "loss": 0.0171, "step": 53350 }, { "grad_norm": 0.2837332487106323, "learning_rate": 4.857986043811836e-05, "loss": 0.0203, "step": 53360 }, { "grad_norm": 0.16005679965019226, "learning_rate": 4.856333248866448e-05, "loss": 0.0168, "step": 53370 }, { "grad_norm": 0.24170777201652527, "learning_rate": 4.85468046963224e-05, "loss": 0.0247, "step": 53380 }, { "grad_norm": 0.25283539295196533, "learning_rate": 4.853027706289957e-05, "loss": 0.0169, "step": 53390 }, { "grad_norm": 0.21034595370292664, "learning_rate": 4.851374959020344e-05, "loss": 0.0128, "step": 53400 }, { "grad_norm": 0.29732969403266907, "learning_rate": 4.849722228004141e-05, "loss": 0.0293, "step": 53410 }, { "grad_norm": 0.26510968804359436, "learning_rate": 4.848069513422091e-05, "loss": 0.0188, "step": 53420 }, { "grad_norm": 0.2262967973947525, "learning_rate": 4.8464168154549294e-05, "loss": 0.0176, "step": 53430 }, { "grad_norm": 0.21664367616176605, "learning_rate": 4.844764134283392e-05, "loss": 0.0167, "step": 53440 }, { "grad_norm": 0.16955137252807617, "learning_rate": 4.843111470088218e-05, "loss": 0.0204, "step": 53450 }, { "grad_norm": 0.24070081114768982, "learning_rate": 4.841458823050137e-05, "loss": 0.0219, "step": 53460 }, { "grad_norm": 0.227405846118927, "learning_rate": 4.8398061933498816e-05, "loss": 0.018, "step": 53470 }, { "grad_norm": 0.3613559305667877, "learning_rate": 4.83815358116818e-05, "loss": 0.019, "step": 53480 }, { "grad_norm": 0.1600925773382187, "learning_rate": 4.8365009866857584e-05, "loss": 0.0167, "step": 53490 }, { "grad_norm": 0.2997353672981262, "learning_rate": 4.834848410083342e-05, "loss": 0.0143, "step": 53500 }, { "grad_norm": 0.28619295358657837, "learning_rate": 4.833195851541657e-05, "loss": 0.0217, "step": 53510 }, { "grad_norm": 0.2538376450538635, "learning_rate": 4.831543311241423e-05, "loss": 0.0171, "step": 53520 }, { "grad_norm": 0.4545162618160248, "learning_rate": 4.829890789363359e-05, "loss": 0.0177, "step": 53530 }, { "grad_norm": 0.2116684764623642, "learning_rate": 4.828238286088182e-05, "loss": 0.0118, "step": 53540 }, { "grad_norm": 0.18649305403232574, "learning_rate": 4.826585801596609e-05, "loss": 0.0192, "step": 53550 }, { "grad_norm": 0.1382620483636856, "learning_rate": 4.8249333360693486e-05, "loss": 0.0139, "step": 53560 }, { "grad_norm": 0.15541554987430573, "learning_rate": 4.8232808896871185e-05, "loss": 0.0121, "step": 53570 }, { "grad_norm": 0.22977277636528015, "learning_rate": 4.821628462630624e-05, "loss": 0.0158, "step": 53580 }, { "grad_norm": 0.3473302721977234, "learning_rate": 4.8199760550805735e-05, "loss": 0.0184, "step": 53590 }, { "grad_norm": 0.27852123975753784, "learning_rate": 4.818323667217669e-05, "loss": 0.0174, "step": 53600 }, { "grad_norm": 0.29450929164886475, "learning_rate": 4.8166712992226165e-05, "loss": 0.02, "step": 53610 }, { "grad_norm": 0.31777819991111755, "learning_rate": 4.8150189512761154e-05, "loss": 0.0204, "step": 53620 }, { "grad_norm": 0.1534174084663391, "learning_rate": 4.813366623558862e-05, "loss": 0.0181, "step": 53630 }, { "grad_norm": 0.32945337891578674, "learning_rate": 4.811714316251554e-05, "loss": 0.0217, "step": 53640 }, { "grad_norm": 0.15930119156837463, "learning_rate": 4.810062029534886e-05, "loss": 0.0129, "step": 53650 }, { "grad_norm": 0.17765499651432037, "learning_rate": 4.8084097635895505e-05, "loss": 0.0221, "step": 53660 }, { "grad_norm": 0.4080914855003357, "learning_rate": 4.8067575185962335e-05, "loss": 0.0173, "step": 53670 }, { "grad_norm": 0.11458656191825867, "learning_rate": 4.805105294735625e-05, "loss": 0.0168, "step": 53680 }, { "grad_norm": 0.20086826384067535, "learning_rate": 4.803453092188406e-05, "loss": 0.0252, "step": 53690 }, { "grad_norm": 0.3333476185798645, "learning_rate": 4.801800911135263e-05, "loss": 0.0249, "step": 53700 }, { "grad_norm": 0.23649118840694427, "learning_rate": 4.8001487517568744e-05, "loss": 0.0136, "step": 53710 }, { "grad_norm": 0.16647674143314362, "learning_rate": 4.798496614233918e-05, "loss": 0.0208, "step": 53720 }, { "grad_norm": 0.15312862396240234, "learning_rate": 4.796844498747069e-05, "loss": 0.0254, "step": 53730 }, { "grad_norm": 0.11822031438350677, "learning_rate": 4.795192405477e-05, "loss": 0.0172, "step": 53740 }, { "grad_norm": 0.36380860209465027, "learning_rate": 4.79354033460438e-05, "loss": 0.0213, "step": 53750 }, { "grad_norm": 0.18596374988555908, "learning_rate": 4.79188828630988e-05, "loss": 0.0239, "step": 53760 }, { "grad_norm": 0.20803308486938477, "learning_rate": 4.790236260774165e-05, "loss": 0.0141, "step": 53770 }, { "grad_norm": 0.17001426219940186, "learning_rate": 4.788584258177896e-05, "loss": 0.0188, "step": 53780 }, { "grad_norm": 0.14319096505641937, "learning_rate": 4.7869322787017356e-05, "loss": 0.0159, "step": 53790 }, { "grad_norm": 0.26941072940826416, "learning_rate": 4.78528032252634e-05, "loss": 0.0194, "step": 53800 }, { "grad_norm": 0.17417393624782562, "learning_rate": 4.783628389832366e-05, "loss": 0.0136, "step": 53810 }, { "grad_norm": 0.17989899218082428, "learning_rate": 4.7819764808004644e-05, "loss": 0.0171, "step": 53820 }, { "grad_norm": 0.10993422567844391, "learning_rate": 4.780324595611289e-05, "loss": 0.0161, "step": 53830 }, { "grad_norm": 0.13288448750972748, "learning_rate": 4.7786727344454854e-05, "loss": 0.0209, "step": 53840 }, { "grad_norm": 0.13290797173976898, "learning_rate": 4.777020897483697e-05, "loss": 0.0209, "step": 53850 }, { "grad_norm": 0.16666872799396515, "learning_rate": 4.775369084906569e-05, "loss": 0.0167, "step": 53860 }, { "grad_norm": 0.26077884435653687, "learning_rate": 4.773717296894742e-05, "loss": 0.0167, "step": 53870 }, { "grad_norm": 0.1857861578464508, "learning_rate": 4.772065533628848e-05, "loss": 0.0136, "step": 53880 }, { "grad_norm": 0.24670328199863434, "learning_rate": 4.7704137952895265e-05, "loss": 0.0135, "step": 53890 }, { "grad_norm": 0.11364082992076874, "learning_rate": 4.768762082057407e-05, "loss": 0.0098, "step": 53900 }, { "grad_norm": 0.510951578617096, "learning_rate": 4.7671103941131186e-05, "loss": 0.0161, "step": 53910 }, { "grad_norm": 0.3975833058357239, "learning_rate": 4.765458731637287e-05, "loss": 0.0215, "step": 53920 }, { "grad_norm": 0.25241023302078247, "learning_rate": 4.763807094810537e-05, "loss": 0.0244, "step": 53930 }, { "grad_norm": 0.19682711362838745, "learning_rate": 4.762155483813487e-05, "loss": 0.0196, "step": 53940 }, { "grad_norm": 0.31103450059890747, "learning_rate": 4.760503898826754e-05, "loss": 0.0208, "step": 53950 }, { "grad_norm": 0.2716829180717468, "learning_rate": 4.758852340030957e-05, "loss": 0.0219, "step": 53960 }, { "grad_norm": 0.205658957362175, "learning_rate": 4.757200807606705e-05, "loss": 0.0241, "step": 53970 }, { "grad_norm": 0.5742418169975281, "learning_rate": 4.755549301734607e-05, "loss": 0.021, "step": 53980 }, { "grad_norm": 0.21623550355434418, "learning_rate": 4.753897822595271e-05, "loss": 0.0239, "step": 53990 }, { "grad_norm": 0.1248791515827179, "learning_rate": 4.7522463703692994e-05, "loss": 0.0152, "step": 54000 }, { "grad_norm": 0.13626545667648315, "learning_rate": 4.75059494523729e-05, "loss": 0.014, "step": 54010 }, { "grad_norm": 0.43866878747940063, "learning_rate": 4.748943547379843e-05, "loss": 0.0213, "step": 54020 }, { "grad_norm": 0.26667138934135437, "learning_rate": 4.7472921769775535e-05, "loss": 0.0187, "step": 54030 }, { "grad_norm": 0.19469064474105835, "learning_rate": 4.745640834211011e-05, "loss": 0.0215, "step": 54040 }, { "grad_norm": 0.16597355902194977, "learning_rate": 4.7439895192608054e-05, "loss": 0.0213, "step": 54050 }, { "grad_norm": 0.21695423126220703, "learning_rate": 4.7423382323075186e-05, "loss": 0.0163, "step": 54060 }, { "grad_norm": 0.07812593132257462, "learning_rate": 4.740686973531737e-05, "loss": 0.0082, "step": 54070 }, { "grad_norm": 0.10048460960388184, "learning_rate": 4.7390357431140375e-05, "loss": 0.0147, "step": 54080 }, { "grad_norm": 0.21368591487407684, "learning_rate": 4.737384541234998e-05, "loss": 0.0166, "step": 54090 }, { "grad_norm": 0.15405234694480896, "learning_rate": 4.735733368075189e-05, "loss": 0.0131, "step": 54100 }, { "grad_norm": 0.16832509636878967, "learning_rate": 4.734082223815182e-05, "loss": 0.0204, "step": 54110 }, { "grad_norm": 0.1398460417985916, "learning_rate": 4.7324311086355436e-05, "loss": 0.0147, "step": 54120 }, { "grad_norm": 0.3292178809642792, "learning_rate": 4.730780022716836e-05, "loss": 0.0205, "step": 54130 }, { "grad_norm": 0.23130953311920166, "learning_rate": 4.729128966239619e-05, "loss": 0.0208, "step": 54140 }, { "grad_norm": 0.26495277881622314, "learning_rate": 4.7274779393844536e-05, "loss": 0.0174, "step": 54150 }, { "grad_norm": 0.27556851506233215, "learning_rate": 4.725826942331891e-05, "loss": 0.0303, "step": 54160 }, { "grad_norm": 0.20676754415035248, "learning_rate": 4.724175975262481e-05, "loss": 0.015, "step": 54170 }, { "grad_norm": 0.24372157454490662, "learning_rate": 4.722525038356772e-05, "loss": 0.0161, "step": 54180 }, { "grad_norm": 0.22847364842891693, "learning_rate": 4.7208741317953075e-05, "loss": 0.0151, "step": 54190 }, { "grad_norm": 0.19546234607696533, "learning_rate": 4.7192232557586264e-05, "loss": 0.0201, "step": 54200 }, { "grad_norm": 0.1406700611114502, "learning_rate": 4.717572410427269e-05, "loss": 0.0237, "step": 54210 }, { "grad_norm": 0.12176458537578583, "learning_rate": 4.71592159598177e-05, "loss": 0.0161, "step": 54220 }, { "grad_norm": 0.14698658883571625, "learning_rate": 4.714270812602657e-05, "loss": 0.0121, "step": 54230 }, { "grad_norm": 0.0980721041560173, "learning_rate": 4.712620060470458e-05, "loss": 0.0294, "step": 54240 }, { "grad_norm": 0.27854123711586, "learning_rate": 4.7109693397656986e-05, "loss": 0.0191, "step": 54250 }, { "grad_norm": 0.3136853277683258, "learning_rate": 4.709318650668894e-05, "loss": 0.0241, "step": 54260 }, { "grad_norm": 0.1708267778158188, "learning_rate": 4.7076679933605676e-05, "loss": 0.0153, "step": 54270 }, { "grad_norm": 0.22134464979171753, "learning_rate": 4.70601736802123e-05, "loss": 0.0225, "step": 54280 }, { "grad_norm": 0.26539549231529236, "learning_rate": 4.704366774831391e-05, "loss": 0.0342, "step": 54290 }, { "grad_norm": 0.3573583662509918, "learning_rate": 4.702716213971557e-05, "loss": 0.0154, "step": 54300 }, { "grad_norm": 0.34004655480384827, "learning_rate": 4.70106568562223e-05, "loss": 0.0188, "step": 54310 }, { "grad_norm": 0.17217715084552765, "learning_rate": 4.6994151899639105e-05, "loss": 0.0164, "step": 54320 }, { "grad_norm": 0.22512651979923248, "learning_rate": 4.697764727177093e-05, "loss": 0.0293, "step": 54330 }, { "grad_norm": 0.26251623034477234, "learning_rate": 4.696114297442272e-05, "loss": 0.0166, "step": 54340 }, { "grad_norm": 0.23903775215148926, "learning_rate": 4.694463900939933e-05, "loss": 0.0113, "step": 54350 }, { "grad_norm": 0.10864751040935516, "learning_rate": 4.692813537850564e-05, "loss": 0.018, "step": 54360 }, { "grad_norm": 0.24920164048671722, "learning_rate": 4.6911632083546445e-05, "loss": 0.017, "step": 54370 }, { "grad_norm": 0.2245679348707199, "learning_rate": 4.6895129126326525e-05, "loss": 0.024, "step": 54380 }, { "grad_norm": 0.12541085481643677, "learning_rate": 4.687862650865059e-05, "loss": 0.0195, "step": 54390 }, { "grad_norm": 0.4386398494243622, "learning_rate": 4.686212423232339e-05, "loss": 0.0241, "step": 54400 }, { "grad_norm": 0.12671077251434326, "learning_rate": 4.6845622299149565e-05, "loss": 0.018, "step": 54410 }, { "grad_norm": 0.18290075659751892, "learning_rate": 4.682912071093374e-05, "loss": 0.015, "step": 54420 }, { "grad_norm": 0.129110649228096, "learning_rate": 4.681261946948052e-05, "loss": 0.0102, "step": 54430 }, { "grad_norm": 0.5202292203903198, "learning_rate": 4.679611857659443e-05, "loss": 0.0209, "step": 54440 }, { "grad_norm": 0.32849639654159546, "learning_rate": 4.6779618034079975e-05, "loss": 0.0139, "step": 54450 }, { "grad_norm": 0.159137561917305, "learning_rate": 4.6763117843741675e-05, "loss": 0.0158, "step": 54460 }, { "grad_norm": 0.1222420409321785, "learning_rate": 4.674661800738394e-05, "loss": 0.0136, "step": 54470 }, { "grad_norm": 0.42055267095565796, "learning_rate": 4.673011852681117e-05, "loss": 0.0132, "step": 54480 }, { "grad_norm": 0.3200419247150421, "learning_rate": 4.671361940382772e-05, "loss": 0.0191, "step": 54490 }, { "grad_norm": 0.1668063849210739, "learning_rate": 4.669712064023791e-05, "loss": 0.0185, "step": 54500 }, { "grad_norm": 0.21486477553844452, "learning_rate": 4.668062223784602e-05, "loss": 0.0145, "step": 54510 }, { "grad_norm": 0.2079809308052063, "learning_rate": 4.6664124198456274e-05, "loss": 0.0181, "step": 54520 }, { "grad_norm": 0.22800475358963013, "learning_rate": 4.6647626523872905e-05, "loss": 0.0209, "step": 54530 }, { "grad_norm": 0.18563298881053925, "learning_rate": 4.6631129215900066e-05, "loss": 0.0118, "step": 54540 }, { "grad_norm": 0.19350387156009674, "learning_rate": 4.661463227634187e-05, "loss": 0.035, "step": 54550 }, { "grad_norm": 0.27470824122428894, "learning_rate": 4.659813570700238e-05, "loss": 0.0118, "step": 54560 }, { "grad_norm": 0.32688701152801514, "learning_rate": 4.658163950968568e-05, "loss": 0.0166, "step": 54570 }, { "grad_norm": 0.08091634511947632, "learning_rate": 4.6565143686195726e-05, "loss": 0.018, "step": 54580 }, { "grad_norm": 0.13527636229991913, "learning_rate": 4.654864823833649e-05, "loss": 0.0205, "step": 54590 }, { "grad_norm": 0.5702047348022461, "learning_rate": 4.65321531679119e-05, "loss": 0.0245, "step": 54600 }, { "grad_norm": 0.13429202139377594, "learning_rate": 4.6515658476725834e-05, "loss": 0.0217, "step": 54610 }, { "grad_norm": 0.3037596642971039, "learning_rate": 4.649916416658212e-05, "loss": 0.0192, "step": 54620 }, { "grad_norm": 0.13536308705806732, "learning_rate": 4.648267023928455e-05, "loss": 0.021, "step": 54630 }, { "grad_norm": 0.13975480198860168, "learning_rate": 4.646617669663687e-05, "loss": 0.0193, "step": 54640 }, { "grad_norm": 0.23296190798282623, "learning_rate": 4.644968354044277e-05, "loss": 0.0204, "step": 54650 }, { "grad_norm": 0.24399256706237793, "learning_rate": 4.643319077250596e-05, "loss": 0.0163, "step": 54660 }, { "grad_norm": 0.21933497488498688, "learning_rate": 4.6416698394630055e-05, "loss": 0.0114, "step": 54670 }, { "grad_norm": 0.07892826944589615, "learning_rate": 4.640020640861862e-05, "loss": 0.0145, "step": 54680 }, { "grad_norm": 0.23183651268482208, "learning_rate": 4.6383714816275195e-05, "loss": 0.0185, "step": 54690 }, { "grad_norm": 0.14493626356124878, "learning_rate": 4.636722361940329e-05, "loss": 0.0167, "step": 54700 }, { "grad_norm": 0.14121796190738678, "learning_rate": 4.635073281980632e-05, "loss": 0.0208, "step": 54710 }, { "grad_norm": 0.2599448561668396, "learning_rate": 4.633424241928775e-05, "loss": 0.0188, "step": 54720 }, { "grad_norm": 0.1704886257648468, "learning_rate": 4.6317752419650914e-05, "loss": 0.0161, "step": 54730 }, { "grad_norm": 0.18158327043056488, "learning_rate": 4.630126282269914e-05, "loss": 0.0154, "step": 54740 }, { "grad_norm": 0.10545884072780609, "learning_rate": 4.628477363023569e-05, "loss": 0.0161, "step": 54750 }, { "grad_norm": 0.21880348026752472, "learning_rate": 4.626828484406383e-05, "loss": 0.0174, "step": 54760 }, { "grad_norm": 0.20649899542331696, "learning_rate": 4.625179646598669e-05, "loss": 0.0277, "step": 54770 }, { "grad_norm": 0.13557197153568268, "learning_rate": 4.6235308497807487e-05, "loss": 0.0179, "step": 54780 }, { "grad_norm": 0.22408337891101837, "learning_rate": 4.6218820941329276e-05, "loss": 0.0168, "step": 54790 }, { "grad_norm": 0.1333593726158142, "learning_rate": 4.620233379835513e-05, "loss": 0.0132, "step": 54800 }, { "grad_norm": 0.17053166031837463, "learning_rate": 4.618584707068804e-05, "loss": 0.0165, "step": 54810 }, { "grad_norm": 0.30204012989997864, "learning_rate": 4.616936076013099e-05, "loss": 0.0148, "step": 54820 }, { "grad_norm": 0.19095812737941742, "learning_rate": 4.615287486848688e-05, "loss": 0.0285, "step": 54830 }, { "grad_norm": 0.14255854487419128, "learning_rate": 4.6136389397558584e-05, "loss": 0.013, "step": 54840 }, { "grad_norm": 0.14498339593410492, "learning_rate": 4.6119904349148935e-05, "loss": 0.0116, "step": 54850 }, { "grad_norm": 0.0874568298459053, "learning_rate": 4.6103419725060726e-05, "loss": 0.0202, "step": 54860 }, { "grad_norm": 0.29223933815956116, "learning_rate": 4.608693552709667e-05, "loss": 0.0171, "step": 54870 }, { "grad_norm": 0.11974120885133743, "learning_rate": 4.607045175705947e-05, "loss": 0.0209, "step": 54880 }, { "grad_norm": 0.3209632337093353, "learning_rate": 4.605396841675176e-05, "loss": 0.0158, "step": 54890 }, { "grad_norm": 0.22395311295986176, "learning_rate": 4.6037485507976094e-05, "loss": 0.0186, "step": 54900 }, { "grad_norm": 0.24192720651626587, "learning_rate": 4.602100303253509e-05, "loss": 0.0158, "step": 54910 }, { "grad_norm": 0.15739476680755615, "learning_rate": 4.600452099223122e-05, "loss": 0.0211, "step": 54920 }, { "grad_norm": 0.1664993017911911, "learning_rate": 4.598803938886691e-05, "loss": 0.0134, "step": 54930 }, { "grad_norm": 0.27585363388061523, "learning_rate": 4.597155822424459e-05, "loss": 0.015, "step": 54940 }, { "grad_norm": 0.4059284031391144, "learning_rate": 4.59550775001666e-05, "loss": 0.0177, "step": 54950 }, { "grad_norm": 0.44346651434898376, "learning_rate": 4.5938597218435234e-05, "loss": 0.023, "step": 54960 }, { "grad_norm": 0.19279834628105164, "learning_rate": 4.592211738085279e-05, "loss": 0.0154, "step": 54970 }, { "grad_norm": 0.12488895654678345, "learning_rate": 4.590563798922146e-05, "loss": 0.0137, "step": 54980 }, { "grad_norm": 0.13439330458641052, "learning_rate": 4.5889159045343404e-05, "loss": 0.0229, "step": 54990 }, { "grad_norm": 0.14124265313148499, "learning_rate": 4.5872680551020734e-05, "loss": 0.0177, "step": 55000 }, { "grad_norm": 0.1578875333070755, "learning_rate": 4.585620250805552e-05, "loss": 0.0166, "step": 55010 }, { "grad_norm": 0.11818012595176697, "learning_rate": 4.583972491824975e-05, "loss": 0.01, "step": 55020 }, { "grad_norm": 0.5068319439888, "learning_rate": 4.582324778340541e-05, "loss": 0.0134, "step": 55030 }, { "grad_norm": 0.21161401271820068, "learning_rate": 4.5806771105324424e-05, "loss": 0.0172, "step": 55040 }, { "grad_norm": 0.3535127341747284, "learning_rate": 4.579029488580865e-05, "loss": 0.0162, "step": 55050 }, { "grad_norm": 0.0927010178565979, "learning_rate": 4.5773819126659876e-05, "loss": 0.0132, "step": 55060 }, { "grad_norm": 0.236318901181221, "learning_rate": 4.575734382967991e-05, "loss": 0.017, "step": 55070 }, { "grad_norm": 0.2089645266532898, "learning_rate": 4.574086899667044e-05, "loss": 0.0107, "step": 55080 }, { "grad_norm": 0.3954615890979767, "learning_rate": 4.5724394629433106e-05, "loss": 0.0158, "step": 55090 }, { "grad_norm": 0.2329573780298233, "learning_rate": 4.570792072976958e-05, "loss": 0.0155, "step": 55100 }, { "grad_norm": 0.2527877688407898, "learning_rate": 4.569144729948138e-05, "loss": 0.0178, "step": 55110 }, { "grad_norm": 0.11051498353481293, "learning_rate": 4.5674974340370033e-05, "loss": 0.0115, "step": 55120 }, { "grad_norm": 0.43045952916145325, "learning_rate": 4.565850185423699e-05, "loss": 0.0231, "step": 55130 }, { "grad_norm": 0.19424059987068176, "learning_rate": 4.564202984288365e-05, "loss": 0.0145, "step": 55140 }, { "grad_norm": 0.45845651626586914, "learning_rate": 4.5625558308111356e-05, "loss": 0.0189, "step": 55150 }, { "grad_norm": 0.28174299001693726, "learning_rate": 4.560908725172145e-05, "loss": 0.0238, "step": 55160 }, { "grad_norm": 0.3213995397090912, "learning_rate": 4.559261667551516e-05, "loss": 0.0117, "step": 55170 }, { "grad_norm": 0.225389301776886, "learning_rate": 4.5576146581293685e-05, "loss": 0.0246, "step": 55180 }, { "grad_norm": 0.14173446595668793, "learning_rate": 4.5559676970858164e-05, "loss": 0.0197, "step": 55190 }, { "grad_norm": 0.5017144083976746, "learning_rate": 4.5543207846009705e-05, "loss": 0.0223, "step": 55200 }, { "grad_norm": 0.20625779032707214, "learning_rate": 4.552673920854933e-05, "loss": 0.0142, "step": 55210 }, { "grad_norm": 0.19290675222873688, "learning_rate": 4.5510271060278006e-05, "loss": 0.0171, "step": 55220 }, { "grad_norm": 0.2808860242366791, "learning_rate": 4.5493803402996704e-05, "loss": 0.0272, "step": 55230 }, { "grad_norm": 0.30071452260017395, "learning_rate": 4.5477336238506295e-05, "loss": 0.0176, "step": 55240 }, { "grad_norm": 0.11815330386161804, "learning_rate": 4.546086956860759e-05, "loss": 0.0101, "step": 55250 }, { "grad_norm": 0.08103864639997482, "learning_rate": 4.5444403395101374e-05, "loss": 0.0159, "step": 55260 }, { "grad_norm": 0.2648201286792755, "learning_rate": 4.542793771978834e-05, "loss": 0.0254, "step": 55270 }, { "grad_norm": 0.27497008442878723, "learning_rate": 4.541147254446916e-05, "loss": 0.0222, "step": 55280 }, { "grad_norm": 0.14403079450130463, "learning_rate": 4.539500787094445e-05, "loss": 0.0158, "step": 55290 }, { "grad_norm": 0.16867317259311676, "learning_rate": 4.537854370101475e-05, "loss": 0.0142, "step": 55300 }, { "grad_norm": 0.09611088782548904, "learning_rate": 4.5362080036480566e-05, "loss": 0.0145, "step": 55310 }, { "grad_norm": 0.2588499188423157, "learning_rate": 4.534561687914235e-05, "loss": 0.0148, "step": 55320 }, { "grad_norm": 0.16836391389369965, "learning_rate": 4.5329154230800466e-05, "loss": 0.0119, "step": 55330 }, { "grad_norm": 0.24921004474163055, "learning_rate": 4.531269209325525e-05, "loss": 0.0221, "step": 55340 }, { "grad_norm": 0.22734102606773376, "learning_rate": 4.529623046830696e-05, "loss": 0.0171, "step": 55350 }, { "grad_norm": 0.217253178358078, "learning_rate": 4.5279769357755866e-05, "loss": 0.0139, "step": 55360 }, { "grad_norm": 0.8650572299957275, "learning_rate": 4.5263308763402084e-05, "loss": 0.0168, "step": 55370 }, { "grad_norm": 0.27817389369010925, "learning_rate": 4.524684868704574e-05, "loss": 0.021, "step": 55380 }, { "grad_norm": 0.19800642132759094, "learning_rate": 4.523038913048688e-05, "loss": 0.0183, "step": 55390 }, { "grad_norm": 0.25922560691833496, "learning_rate": 4.521393009552549e-05, "loss": 0.019, "step": 55400 }, { "grad_norm": 0.35520291328430176, "learning_rate": 4.5197471583961484e-05, "loss": 0.0215, "step": 55410 }, { "grad_norm": 0.2746155560016632, "learning_rate": 4.51810135975948e-05, "loss": 0.0247, "step": 55420 }, { "grad_norm": 0.12106389552354813, "learning_rate": 4.516455613822521e-05, "loss": 0.0113, "step": 55430 }, { "grad_norm": 0.18702244758605957, "learning_rate": 4.514809920765249e-05, "loss": 0.0175, "step": 55440 }, { "grad_norm": 0.4407242238521576, "learning_rate": 4.5131642807676356e-05, "loss": 0.015, "step": 55450 }, { "grad_norm": 0.39056462049484253, "learning_rate": 4.511518694009644e-05, "loss": 0.0139, "step": 55460 }, { "grad_norm": 0.20110206305980682, "learning_rate": 4.509873160671231e-05, "loss": 0.0202, "step": 55470 }, { "grad_norm": 0.365233838558197, "learning_rate": 4.508227680932356e-05, "loss": 0.0185, "step": 55480 }, { "grad_norm": 0.13830161094665527, "learning_rate": 4.5065822549729605e-05, "loss": 0.0195, "step": 55490 }, { "grad_norm": 0.25130242109298706, "learning_rate": 4.504936882972989e-05, "loss": 0.0146, "step": 55500 }, { "grad_norm": 0.25376203656196594, "learning_rate": 4.503291565112374e-05, "loss": 0.0107, "step": 55510 }, { "grad_norm": 0.2759241461753845, "learning_rate": 4.501646301571048e-05, "loss": 0.0154, "step": 55520 }, { "grad_norm": 0.18804065883159637, "learning_rate": 4.500001092528933e-05, "loss": 0.0135, "step": 55530 }, { "grad_norm": 0.21762573719024658, "learning_rate": 4.4983559381659455e-05, "loss": 0.0168, "step": 55540 }, { "grad_norm": 0.42026251554489136, "learning_rate": 4.496710838661999e-05, "loss": 0.0215, "step": 55550 }, { "grad_norm": 0.14815865457057953, "learning_rate": 4.495065794196999e-05, "loss": 0.0181, "step": 55560 }, { "grad_norm": 0.2433955818414688, "learning_rate": 4.4934208049508455e-05, "loss": 0.0142, "step": 55570 }, { "grad_norm": 0.24812975525856018, "learning_rate": 4.4917758711034303e-05, "loss": 0.0214, "step": 55580 }, { "grad_norm": 0.17505285143852234, "learning_rate": 4.490130992834642e-05, "loss": 0.0163, "step": 55590 }, { "grad_norm": 0.1552920788526535, "learning_rate": 4.488486170324359e-05, "loss": 0.0169, "step": 55600 }, { "grad_norm": 0.13006065785884857, "learning_rate": 4.486841403752463e-05, "loss": 0.0186, "step": 55610 }, { "grad_norm": 0.21298062801361084, "learning_rate": 4.4851966932988174e-05, "loss": 0.0175, "step": 55620 }, { "grad_norm": 0.2262909710407257, "learning_rate": 4.4835520391432894e-05, "loss": 0.0126, "step": 55630 }, { "grad_norm": 0.3652310073375702, "learning_rate": 4.481907441465733e-05, "loss": 0.0241, "step": 55640 }, { "grad_norm": 0.21556927263736725, "learning_rate": 4.480262900446e-05, "loss": 0.0149, "step": 55650 }, { "grad_norm": 0.1750839501619339, "learning_rate": 4.478618416263933e-05, "loss": 0.0144, "step": 55660 }, { "grad_norm": 0.45965197682380676, "learning_rate": 4.476973989099373e-05, "loss": 0.0199, "step": 55670 }, { "grad_norm": 0.4564456343650818, "learning_rate": 4.475329619132153e-05, "loss": 0.0234, "step": 55680 }, { "grad_norm": 0.20244404673576355, "learning_rate": 4.4736853065420956e-05, "loss": 0.0177, "step": 55690 }, { "grad_norm": 0.4227379560470581, "learning_rate": 4.472041051509023e-05, "loss": 0.0162, "step": 55700 }, { "grad_norm": 0.19888480007648468, "learning_rate": 4.470396854212746e-05, "loss": 0.0126, "step": 55710 }, { "grad_norm": 0.06289824843406677, "learning_rate": 4.468752714833072e-05, "loss": 0.014, "step": 55720 }, { "grad_norm": 0.30361658334732056, "learning_rate": 4.467108633549803e-05, "loss": 0.011, "step": 55730 }, { "grad_norm": 0.26762405037879944, "learning_rate": 4.4654646105427335e-05, "loss": 0.0281, "step": 55740 }, { "grad_norm": 0.18158148229122162, "learning_rate": 4.463820645991651e-05, "loss": 0.0209, "step": 55750 }, { "grad_norm": 0.25591641664505005, "learning_rate": 4.462176740076334e-05, "loss": 0.0167, "step": 55760 }, { "grad_norm": 0.12249033898115158, "learning_rate": 4.4605328929765624e-05, "loss": 0.016, "step": 55770 }, { "grad_norm": 0.31087133288383484, "learning_rate": 4.458889104872102e-05, "loss": 0.0128, "step": 55780 }, { "grad_norm": 0.24763154983520508, "learning_rate": 4.457245375942715e-05, "loss": 0.0209, "step": 55790 }, { "grad_norm": 0.19172927737236023, "learning_rate": 4.455601706368157e-05, "loss": 0.0182, "step": 55800 }, { "grad_norm": 0.12089163810014725, "learning_rate": 4.453958096328179e-05, "loss": 0.022, "step": 55810 }, { "grad_norm": 0.24928410351276398, "learning_rate": 4.452314546002523e-05, "loss": 0.0221, "step": 55820 }, { "grad_norm": 0.28895220160484314, "learning_rate": 4.4506710555709246e-05, "loss": 0.0112, "step": 55830 }, { "grad_norm": 0.14654244482517242, "learning_rate": 4.449027625213114e-05, "loss": 0.0201, "step": 55840 }, { "grad_norm": 0.3248876929283142, "learning_rate": 4.4473842551088135e-05, "loss": 0.0162, "step": 55850 }, { "grad_norm": 0.2948611378669739, "learning_rate": 4.445740945437738e-05, "loss": 0.0152, "step": 55860 }, { "grad_norm": 0.33977410197257996, "learning_rate": 4.444097696379602e-05, "loss": 0.0226, "step": 55870 }, { "grad_norm": 0.29732972383499146, "learning_rate": 4.442454508114106e-05, "loss": 0.0266, "step": 55880 }, { "grad_norm": 0.1664678007364273, "learning_rate": 4.4408113808209465e-05, "loss": 0.0171, "step": 55890 }, { "grad_norm": 0.1438388079404831, "learning_rate": 4.439168314679813e-05, "loss": 0.0132, "step": 55900 }, { "grad_norm": 0.18560992181301117, "learning_rate": 4.43752530987039e-05, "loss": 0.0149, "step": 55910 }, { "grad_norm": 0.23406793177127838, "learning_rate": 4.4358823665723515e-05, "loss": 0.0084, "step": 55920 }, { "grad_norm": 0.15958645939826965, "learning_rate": 4.434239484965371e-05, "loss": 0.0132, "step": 55930 }, { "grad_norm": 0.17000801861286163, "learning_rate": 4.4325966652291103e-05, "loss": 0.0195, "step": 55940 }, { "grad_norm": 0.17212454974651337, "learning_rate": 4.430953907543225e-05, "loss": 0.0229, "step": 55950 }, { "grad_norm": 0.0996011272072792, "learning_rate": 4.4293112120873645e-05, "loss": 0.0204, "step": 55960 }, { "grad_norm": 0.19333341717720032, "learning_rate": 4.427668579041172e-05, "loss": 0.0098, "step": 55970 }, { "grad_norm": 0.13726049661636353, "learning_rate": 4.4260260085842816e-05, "loss": 0.0199, "step": 55980 }, { "grad_norm": 0.14660266041755676, "learning_rate": 4.4243835008963256e-05, "loss": 0.0133, "step": 55990 }, { "grad_norm": 0.12270785123109818, "learning_rate": 4.422741056156925e-05, "loss": 0.0078, "step": 56000 }, { "grad_norm": 0.2479795515537262, "learning_rate": 4.421098674545693e-05, "loss": 0.0107, "step": 56010 }, { "grad_norm": 0.1868908405303955, "learning_rate": 4.419456356242241e-05, "loss": 0.016, "step": 56020 }, { "grad_norm": 0.15386995673179626, "learning_rate": 4.4178141014261685e-05, "loss": 0.0161, "step": 56030 }, { "grad_norm": 0.23114559054374695, "learning_rate": 4.416171910277071e-05, "loss": 0.0137, "step": 56040 }, { "grad_norm": 0.2571868598461151, "learning_rate": 4.414529782974533e-05, "loss": 0.0138, "step": 56050 }, { "grad_norm": 0.08849186450242996, "learning_rate": 4.41288771969814e-05, "loss": 0.0142, "step": 56060 }, { "grad_norm": 0.26913726329803467, "learning_rate": 4.411245720627464e-05, "loss": 0.013, "step": 56070 }, { "grad_norm": 0.35433778166770935, "learning_rate": 4.40960378594207e-05, "loss": 0.0198, "step": 56080 }, { "grad_norm": 0.16171002388000488, "learning_rate": 4.407961915821519e-05, "loss": 0.016, "step": 56090 }, { "grad_norm": 0.26379427313804626, "learning_rate": 4.406320110445362e-05, "loss": 0.0199, "step": 56100 }, { "grad_norm": 0.23387424647808075, "learning_rate": 4.404678369993144e-05, "loss": 0.0142, "step": 56110 }, { "grad_norm": 0.22411945462226868, "learning_rate": 4.403036694644406e-05, "loss": 0.0169, "step": 56120 }, { "grad_norm": 0.1490607112646103, "learning_rate": 4.4013950845786764e-05, "loss": 0.0155, "step": 56130 }, { "grad_norm": 0.5368213653564453, "learning_rate": 4.399753539975482e-05, "loss": 0.0254, "step": 56140 }, { "grad_norm": 0.2012900561094284, "learning_rate": 4.398112061014337e-05, "loss": 0.0111, "step": 56150 }, { "grad_norm": 0.293984591960907, "learning_rate": 4.396470647874753e-05, "loss": 0.0207, "step": 56160 }, { "grad_norm": 0.1545845866203308, "learning_rate": 4.394829300736229e-05, "loss": 0.0152, "step": 56170 }, { "grad_norm": 0.26680630445480347, "learning_rate": 4.393188019778265e-05, "loss": 0.0164, "step": 56180 }, { "grad_norm": 0.11811108142137527, "learning_rate": 4.391546805180347e-05, "loss": 0.0143, "step": 56190 }, { "grad_norm": 0.23785756528377533, "learning_rate": 4.389905657121955e-05, "loss": 0.0171, "step": 56200 }, { "grad_norm": 0.2887231707572937, "learning_rate": 4.3882645757825635e-05, "loss": 0.0208, "step": 56210 }, { "grad_norm": 0.15704672038555145, "learning_rate": 4.386623561341637e-05, "loss": 0.0153, "step": 56220 }, { "grad_norm": 0.06845678389072418, "learning_rate": 4.384982613978637e-05, "loss": 0.0183, "step": 56230 }, { "grad_norm": 0.14540866017341614, "learning_rate": 4.383341733873012e-05, "loss": 0.0151, "step": 56240 }, { "grad_norm": 0.24918906390666962, "learning_rate": 4.381700921204209e-05, "loss": 0.0152, "step": 56250 }, { "grad_norm": 0.4564264118671417, "learning_rate": 4.380060176151663e-05, "loss": 0.0226, "step": 56260 }, { "grad_norm": 0.4276781678199768, "learning_rate": 4.378419498894805e-05, "loss": 0.0196, "step": 56270 }, { "grad_norm": 0.11260350048542023, "learning_rate": 4.376778889613056e-05, "loss": 0.0166, "step": 56280 }, { "grad_norm": 0.11592064797878265, "learning_rate": 4.375138348485831e-05, "loss": 0.0196, "step": 56290 }, { "grad_norm": 0.22519226372241974, "learning_rate": 4.3734978756925346e-05, "loss": 0.0162, "step": 56300 }, { "grad_norm": 0.18481630086898804, "learning_rate": 4.3718574714125706e-05, "loss": 0.0131, "step": 56310 }, { "grad_norm": 0.25763899087905884, "learning_rate": 4.370217135825329e-05, "loss": 0.0102, "step": 56320 }, { "grad_norm": 0.14172226190567017, "learning_rate": 4.368576869110194e-05, "loss": 0.013, "step": 56330 }, { "grad_norm": 0.2840147316455841, "learning_rate": 4.366936671446544e-05, "loss": 0.0202, "step": 56340 }, { "grad_norm": 0.08780808746814728, "learning_rate": 4.3652965430137474e-05, "loss": 0.0215, "step": 56350 }, { "grad_norm": 0.12950348854064941, "learning_rate": 4.3636564839911646e-05, "loss": 0.0108, "step": 56360 }, { "grad_norm": 0.12676915526390076, "learning_rate": 4.3620164945581545e-05, "loss": 0.0113, "step": 56370 }, { "grad_norm": 0.19804394245147705, "learning_rate": 4.360376574894061e-05, "loss": 0.0188, "step": 56380 }, { "grad_norm": 0.22590722143650055, "learning_rate": 4.358736725178224e-05, "loss": 0.0145, "step": 56390 }, { "grad_norm": 0.4082168936729431, "learning_rate": 4.357096945589974e-05, "loss": 0.0238, "step": 56400 }, { "grad_norm": 0.16355368494987488, "learning_rate": 4.355457236308636e-05, "loss": 0.0129, "step": 56410 }, { "grad_norm": 0.23372085392475128, "learning_rate": 4.353817597513526e-05, "loss": 0.0116, "step": 56420 }, { "grad_norm": 0.12787888944149017, "learning_rate": 4.352178029383948e-05, "loss": 0.0154, "step": 56430 }, { "grad_norm": 0.27693644165992737, "learning_rate": 4.3505385320992105e-05, "loss": 0.0157, "step": 56440 }, { "grad_norm": 0.22271022200584412, "learning_rate": 4.348899105838602e-05, "loss": 0.0181, "step": 56450 }, { "grad_norm": 0.229594424366951, "learning_rate": 4.3472597507814087e-05, "loss": 0.0208, "step": 56460 }, { "grad_norm": 0.2308540642261505, "learning_rate": 4.3456204671069066e-05, "loss": 0.0209, "step": 56470 }, { "grad_norm": 0.24846188724040985, "learning_rate": 4.343981254994367e-05, "loss": 0.015, "step": 56480 }, { "grad_norm": 0.20831650495529175, "learning_rate": 4.34234211462305e-05, "loss": 0.0222, "step": 56490 }, { "grad_norm": 0.39954695105552673, "learning_rate": 4.340703046172213e-05, "loss": 0.0165, "step": 56500 }, { "grad_norm": 0.09387248009443283, "learning_rate": 4.339064049821097e-05, "loss": 0.0115, "step": 56510 }, { "grad_norm": 0.3803008496761322, "learning_rate": 4.3374251257489446e-05, "loss": 0.0178, "step": 56520 }, { "grad_norm": 0.1518404632806778, "learning_rate": 4.335786274134984e-05, "loss": 0.0197, "step": 56530 }, { "grad_norm": 0.16382941603660583, "learning_rate": 4.334147495158438e-05, "loss": 0.0249, "step": 56540 }, { "grad_norm": 0.6426345109939575, "learning_rate": 4.33250878899852e-05, "loss": 0.0217, "step": 56550 }, { "grad_norm": 0.19011342525482178, "learning_rate": 4.330870155834437e-05, "loss": 0.0204, "step": 56560 }, { "grad_norm": 0.26157689094543457, "learning_rate": 4.3292315958453884e-05, "loss": 0.0128, "step": 56570 }, { "grad_norm": 0.22178475558757782, "learning_rate": 4.3275931092105646e-05, "loss": 0.0158, "step": 56580 }, { "grad_norm": 0.2196454554796219, "learning_rate": 4.325954696109147e-05, "loss": 0.0133, "step": 56590 }, { "grad_norm": 0.30886515974998474, "learning_rate": 4.32431635672031e-05, "loss": 0.0199, "step": 56600 }, { "grad_norm": 0.20806819200515747, "learning_rate": 4.3226780912232214e-05, "loss": 0.0221, "step": 56610 }, { "grad_norm": 0.11796154826879501, "learning_rate": 4.321039899797036e-05, "loss": 0.0118, "step": 56620 }, { "grad_norm": 0.15756583213806152, "learning_rate": 4.319401782620908e-05, "loss": 0.0131, "step": 56630 }, { "grad_norm": 0.3519999086856842, "learning_rate": 4.317763739873978e-05, "loss": 0.0211, "step": 56640 }, { "grad_norm": 0.3963201344013214, "learning_rate": 4.31612577173538e-05, "loss": 0.0175, "step": 56650 }, { "grad_norm": 0.27091142535209656, "learning_rate": 4.314487878384239e-05, "loss": 0.0264, "step": 56660 }, { "grad_norm": 0.1347140520811081, "learning_rate": 4.3128500599996736e-05, "loss": 0.014, "step": 56670 }, { "grad_norm": 0.11231555789709091, "learning_rate": 4.31121231676079e-05, "loss": 0.0154, "step": 56680 }, { "grad_norm": 0.23938383162021637, "learning_rate": 4.309574648846694e-05, "loss": 0.0211, "step": 56690 }, { "grad_norm": 0.21615491807460785, "learning_rate": 4.3079370564364755e-05, "loss": 0.0166, "step": 56700 }, { "grad_norm": 0.2854405641555786, "learning_rate": 4.306299539709221e-05, "loss": 0.0108, "step": 56710 }, { "grad_norm": 0.24378328025341034, "learning_rate": 4.3046620988440046e-05, "loss": 0.0175, "step": 56720 }, { "grad_norm": 0.24778206646442413, "learning_rate": 4.303024734019897e-05, "loss": 0.0115, "step": 56730 }, { "grad_norm": 0.1981799155473709, "learning_rate": 4.3013874454159565e-05, "loss": 0.0206, "step": 56740 }, { "grad_norm": 0.15580108761787415, "learning_rate": 4.299750233211233e-05, "loss": 0.0231, "step": 56750 }, { "grad_norm": 0.3050002157688141, "learning_rate": 4.2981130975847715e-05, "loss": 0.0184, "step": 56760 }, { "grad_norm": 0.1420598030090332, "learning_rate": 4.2964760387156076e-05, "loss": 0.0151, "step": 56770 }, { "grad_norm": 0.3287816047668457, "learning_rate": 4.294839056782767e-05, "loss": 0.0157, "step": 56780 }, { "grad_norm": 0.6198928356170654, "learning_rate": 4.293202151965266e-05, "loss": 0.0151, "step": 56790 }, { "grad_norm": 0.1489485204219818, "learning_rate": 4.291565324442115e-05, "loss": 0.0139, "step": 56800 }, { "grad_norm": 0.2626228332519531, "learning_rate": 4.2899285743923135e-05, "loss": 0.0209, "step": 56810 }, { "grad_norm": 0.21599431335926056, "learning_rate": 4.288291901994857e-05, "loss": 0.0176, "step": 56820 }, { "grad_norm": 0.21600614488124847, "learning_rate": 4.286655307428729e-05, "loss": 0.0163, "step": 56830 }, { "grad_norm": 0.18438637256622314, "learning_rate": 4.285018790872903e-05, "loss": 0.0259, "step": 56840 }, { "grad_norm": 0.3395422399044037, "learning_rate": 4.283382352506347e-05, "loss": 0.0166, "step": 56850 }, { "grad_norm": 0.1560499221086502, "learning_rate": 4.2817459925080205e-05, "loss": 0.0231, "step": 56860 }, { "grad_norm": 0.1772109568119049, "learning_rate": 4.28010971105687e-05, "loss": 0.0166, "step": 56870 }, { "grad_norm": 0.13716398179531097, "learning_rate": 4.278473508331841e-05, "loss": 0.0155, "step": 56880 }, { "grad_norm": 0.20094560086727142, "learning_rate": 4.276837384511864e-05, "loss": 0.0212, "step": 56890 }, { "grad_norm": 0.34233835339546204, "learning_rate": 4.275201339775864e-05, "loss": 0.0129, "step": 56900 }, { "grad_norm": 0.2711612582206726, "learning_rate": 4.273565374302756e-05, "loss": 0.0174, "step": 56910 }, { "grad_norm": 0.3452628552913666, "learning_rate": 4.2719294882714454e-05, "loss": 0.0136, "step": 56920 }, { "grad_norm": 0.14709238708019257, "learning_rate": 4.270293681860831e-05, "loss": 0.0247, "step": 56930 }, { "grad_norm": 0.185049369931221, "learning_rate": 4.268657955249802e-05, "loss": 0.0219, "step": 56940 }, { "grad_norm": 0.23037537932395935, "learning_rate": 4.267022308617241e-05, "loss": 0.0132, "step": 56950 }, { "grad_norm": 0.31793448328971863, "learning_rate": 4.265386742142018e-05, "loss": 0.0169, "step": 56960 }, { "grad_norm": 0.17318926751613617, "learning_rate": 4.263751256002996e-05, "loss": 0.016, "step": 56970 }, { "grad_norm": 0.111677385866642, "learning_rate": 4.26211585037903e-05, "loss": 0.0146, "step": 56980 }, { "grad_norm": 0.14129354059696198, "learning_rate": 4.2604805254489656e-05, "loss": 0.0172, "step": 56990 }, { "grad_norm": 0.23380514979362488, "learning_rate": 4.258845281391638e-05, "loss": 0.0152, "step": 57000 }, { "grad_norm": 0.20596717298030853, "learning_rate": 4.257210118385876e-05, "loss": 0.0275, "step": 57010 }, { "grad_norm": 0.3309381306171417, "learning_rate": 4.2555750366104994e-05, "loss": 0.0176, "step": 57020 }, { "grad_norm": 0.23640350997447968, "learning_rate": 4.2539400362443184e-05, "loss": 0.0145, "step": 57030 }, { "grad_norm": 0.21235619485378265, "learning_rate": 4.252305117466133e-05, "loss": 0.0182, "step": 57040 }, { "grad_norm": 0.297434002161026, "learning_rate": 4.2506702804547356e-05, "loss": 0.0296, "step": 57050 }, { "grad_norm": 0.19102539122104645, "learning_rate": 4.249035525388907e-05, "loss": 0.0161, "step": 57060 }, { "grad_norm": 0.22093133628368378, "learning_rate": 4.247400852447427e-05, "loss": 0.0095, "step": 57070 }, { "grad_norm": 0.07740797102451324, "learning_rate": 4.245766261809059e-05, "loss": 0.0132, "step": 57080 }, { "grad_norm": 0.16332432627677917, "learning_rate": 4.244131753652558e-05, "loss": 0.02, "step": 57090 }, { "grad_norm": 0.1316106766462326, "learning_rate": 4.242497328156672e-05, "loss": 0.0167, "step": 57100 }, { "grad_norm": 0.4220943748950958, "learning_rate": 4.240862985500139e-05, "loss": 0.0312, "step": 57110 }, { "grad_norm": 0.14099325239658356, "learning_rate": 4.2392287258616894e-05, "loss": 0.0125, "step": 57120 }, { "grad_norm": 0.1296955645084381, "learning_rate": 4.2375945494200396e-05, "loss": 0.0083, "step": 57130 }, { "grad_norm": 0.3264605402946472, "learning_rate": 4.2359604563539054e-05, "loss": 0.0171, "step": 57140 }, { "grad_norm": 0.2554468512535095, "learning_rate": 4.2343264468419885e-05, "loss": 0.0175, "step": 57150 }, { "grad_norm": 0.1211380809545517, "learning_rate": 4.232692521062979e-05, "loss": 0.0149, "step": 57160 }, { "grad_norm": 0.3298974335193634, "learning_rate": 4.231058679195562e-05, "loss": 0.0182, "step": 57170 }, { "grad_norm": 0.24256864190101624, "learning_rate": 4.229424921418411e-05, "loss": 0.0166, "step": 57180 }, { "grad_norm": 0.3085628151893616, "learning_rate": 4.2277912479101916e-05, "loss": 0.0158, "step": 57190 }, { "grad_norm": 0.28550830483436584, "learning_rate": 4.226157658849562e-05, "loss": 0.0124, "step": 57200 }, { "grad_norm": 0.1521552950143814, "learning_rate": 4.2245241544151664e-05, "loss": 0.0141, "step": 57210 }, { "grad_norm": 0.14151974022388458, "learning_rate": 4.2228907347856425e-05, "loss": 0.0118, "step": 57220 }, { "grad_norm": 0.20333339273929596, "learning_rate": 4.2212574001396214e-05, "loss": 0.0183, "step": 57230 }, { "grad_norm": 0.23017290234565735, "learning_rate": 4.21962415065572e-05, "loss": 0.0178, "step": 57240 }, { "grad_norm": 0.1711849421262741, "learning_rate": 4.217990986512548e-05, "loss": 0.0209, "step": 57250 }, { "grad_norm": 0.10754192620515823, "learning_rate": 4.216357907888704e-05, "loss": 0.0228, "step": 57260 }, { "grad_norm": 0.2507517337799072, "learning_rate": 4.2147249149627824e-05, "loss": 0.0156, "step": 57270 }, { "grad_norm": 0.19856233894824982, "learning_rate": 4.213092007913364e-05, "loss": 0.0168, "step": 57280 }, { "grad_norm": 0.2865222096443176, "learning_rate": 4.2114591869190205e-05, "loss": 0.0284, "step": 57290 }, { "grad_norm": 0.5245824456214905, "learning_rate": 4.209826452158315e-05, "loss": 0.0219, "step": 57300 }, { "grad_norm": 0.5570971965789795, "learning_rate": 4.2081938038098e-05, "loss": 0.0184, "step": 57310 }, { "grad_norm": 0.2137419730424881, "learning_rate": 4.2065612420520185e-05, "loss": 0.0144, "step": 57320 }, { "grad_norm": 0.2842436134815216, "learning_rate": 4.204928767063509e-05, "loss": 0.022, "step": 57330 }, { "grad_norm": 0.2700432538986206, "learning_rate": 4.2032963790227944e-05, "loss": 0.0208, "step": 57340 }, { "grad_norm": 0.25926220417022705, "learning_rate": 4.2016640781083887e-05, "loss": 0.0136, "step": 57350 }, { "grad_norm": 0.3269456624984741, "learning_rate": 4.2000318644988004e-05, "loss": 0.0324, "step": 57360 }, { "grad_norm": 0.11376649886369705, "learning_rate": 4.198399738372524e-05, "loss": 0.0171, "step": 57370 }, { "grad_norm": 0.15868420898914337, "learning_rate": 4.196767699908045e-05, "loss": 0.0152, "step": 57380 }, { "grad_norm": 0.500203013420105, "learning_rate": 4.195135749283845e-05, "loss": 0.0213, "step": 57390 }, { "grad_norm": 0.21702027320861816, "learning_rate": 4.1935038866783894e-05, "loss": 0.0168, "step": 57400 }, { "grad_norm": 0.23477709293365479, "learning_rate": 4.1918721122701355e-05, "loss": 0.0188, "step": 57410 }, { "grad_norm": 0.29649460315704346, "learning_rate": 4.190240426237533e-05, "loss": 0.0269, "step": 57420 }, { "grad_norm": 0.24584440886974335, "learning_rate": 4.188608828759019e-05, "loss": 0.0137, "step": 57430 }, { "grad_norm": 0.21155793964862823, "learning_rate": 4.1869773200130246e-05, "loss": 0.0219, "step": 57440 }, { "grad_norm": 0.19087877869606018, "learning_rate": 4.185345900177967e-05, "loss": 0.0215, "step": 57450 }, { "grad_norm": 0.07627858966588974, "learning_rate": 4.183714569432258e-05, "loss": 0.0157, "step": 57460 }, { "grad_norm": 0.08014621585607529, "learning_rate": 4.182083327954295e-05, "loss": 0.0218, "step": 57470 }, { "grad_norm": 0.24378228187561035, "learning_rate": 4.1804521759224714e-05, "loss": 0.0167, "step": 57480 }, { "grad_norm": 0.17349492013454437, "learning_rate": 4.178821113515165e-05, "loss": 0.0189, "step": 57490 }, { "grad_norm": 0.11763391643762589, "learning_rate": 4.1771901409107476e-05, "loss": 0.0183, "step": 57500 }, { "grad_norm": 0.13183866441249847, "learning_rate": 4.175559258287577e-05, "loss": 0.0219, "step": 57510 }, { "grad_norm": 0.17898903787136078, "learning_rate": 4.173928465824008e-05, "loss": 0.0141, "step": 57520 }, { "grad_norm": 0.33784157037734985, "learning_rate": 4.17229776369838e-05, "loss": 0.0207, "step": 57530 }, { "grad_norm": 0.27256879210472107, "learning_rate": 4.170667152089025e-05, "loss": 0.0196, "step": 57540 }, { "grad_norm": 0.14831343293190002, "learning_rate": 4.1690366311742626e-05, "loss": 0.0098, "step": 57550 }, { "grad_norm": 0.14094103872776031, "learning_rate": 4.1674062011324055e-05, "loss": 0.0083, "step": 57560 }, { "grad_norm": 0.21808049082756042, "learning_rate": 4.165775862141752e-05, "loss": 0.013, "step": 57570 }, { "grad_norm": 0.20113983750343323, "learning_rate": 4.164145614380598e-05, "loss": 0.0156, "step": 57580 }, { "grad_norm": 0.17245392501354218, "learning_rate": 4.162515458027223e-05, "loss": 0.0163, "step": 57590 }, { "grad_norm": 0.21504980325698853, "learning_rate": 4.160885393259899e-05, "loss": 0.0276, "step": 57600 }, { "grad_norm": 0.2706320583820343, "learning_rate": 4.159255420256886e-05, "loss": 0.0196, "step": 57610 }, { "grad_norm": 0.21446678042411804, "learning_rate": 4.157625539196436e-05, "loss": 0.0138, "step": 57620 }, { "grad_norm": 0.20641648769378662, "learning_rate": 4.155995750256791e-05, "loss": 0.0177, "step": 57630 }, { "grad_norm": 0.17316687107086182, "learning_rate": 4.15436605361618e-05, "loss": 0.0132, "step": 57640 }, { "grad_norm": 0.20466962456703186, "learning_rate": 4.152736449452827e-05, "loss": 0.0206, "step": 57650 }, { "grad_norm": 0.30877649784088135, "learning_rate": 4.151106937944943e-05, "loss": 0.0156, "step": 57660 }, { "grad_norm": 0.17361781001091003, "learning_rate": 4.149477519270727e-05, "loss": 0.0166, "step": 57670 }, { "grad_norm": 0.1089918315410614, "learning_rate": 4.147848193608371e-05, "loss": 0.0187, "step": 57680 }, { "grad_norm": 0.13095632195472717, "learning_rate": 4.146218961136056e-05, "loss": 0.0082, "step": 57690 }, { "grad_norm": 0.13781389594078064, "learning_rate": 4.14458982203195e-05, "loss": 0.0126, "step": 57700 }, { "grad_norm": 0.24558329582214355, "learning_rate": 4.1429607764742165e-05, "loss": 0.0114, "step": 57710 }, { "grad_norm": 0.05616763234138489, "learning_rate": 4.141331824641003e-05, "loss": 0.0136, "step": 57720 }, { "grad_norm": 0.3854748010635376, "learning_rate": 4.139702966710452e-05, "loss": 0.015, "step": 57730 }, { "grad_norm": 0.3717804551124573, "learning_rate": 4.13807420286069e-05, "loss": 0.0138, "step": 57740 }, { "grad_norm": 0.15421637892723083, "learning_rate": 4.136445533269837e-05, "loss": 0.013, "step": 57750 }, { "grad_norm": 0.3045820891857147, "learning_rate": 4.134816958116003e-05, "loss": 0.0185, "step": 57760 }, { "grad_norm": 0.291545569896698, "learning_rate": 4.133188477577283e-05, "loss": 0.0207, "step": 57770 }, { "grad_norm": 0.14626048505306244, "learning_rate": 4.131560091831771e-05, "loss": 0.0128, "step": 57780 }, { "grad_norm": 0.059214744716882706, "learning_rate": 4.1299318010575405e-05, "loss": 0.0121, "step": 57790 }, { "grad_norm": 0.1628091186285019, "learning_rate": 4.128303605432661e-05, "loss": 0.0348, "step": 57800 }, { "grad_norm": 0.2709604501724243, "learning_rate": 4.126675505135188e-05, "loss": 0.0127, "step": 57810 }, { "grad_norm": 0.405497282743454, "learning_rate": 4.125047500343169e-05, "loss": 0.019, "step": 57820 }, { "grad_norm": 0.31310948729515076, "learning_rate": 4.1234195912346376e-05, "loss": 0.0133, "step": 57830 }, { "grad_norm": 0.07501577585935593, "learning_rate": 4.1217917779876235e-05, "loss": 0.0142, "step": 57840 }, { "grad_norm": 0.13828200101852417, "learning_rate": 4.120164060780141e-05, "loss": 0.0184, "step": 57850 }, { "grad_norm": 0.30584385991096497, "learning_rate": 4.1185364397901935e-05, "loss": 0.0118, "step": 57860 }, { "grad_norm": 0.2269441932439804, "learning_rate": 4.116908915195776e-05, "loss": 0.015, "step": 57870 }, { "grad_norm": 0.35045576095581055, "learning_rate": 4.115281487174872e-05, "loss": 0.0273, "step": 57880 }, { "grad_norm": 0.2285887748003006, "learning_rate": 4.113654155905452e-05, "loss": 0.0108, "step": 57890 }, { "grad_norm": 0.18997322022914886, "learning_rate": 4.112026921565483e-05, "loss": 0.0149, "step": 57900 }, { "grad_norm": 0.25906744599342346, "learning_rate": 4.1103997843329147e-05, "loss": 0.0117, "step": 57910 }, { "grad_norm": 0.2686215043067932, "learning_rate": 4.108772744385689e-05, "loss": 0.0335, "step": 57920 }, { "grad_norm": 0.2491653859615326, "learning_rate": 4.1071458019017346e-05, "loss": 0.018, "step": 57930 }, { "grad_norm": 0.19874753057956696, "learning_rate": 4.1055189570589734e-05, "loss": 0.0201, "step": 57940 }, { "grad_norm": 0.4418153762817383, "learning_rate": 4.103892210035315e-05, "loss": 0.0159, "step": 57950 }, { "grad_norm": 0.3769410252571106, "learning_rate": 4.102265561008656e-05, "loss": 0.0212, "step": 57960 }, { "grad_norm": 0.34459489583969116, "learning_rate": 4.100639010156885e-05, "loss": 0.01, "step": 57970 }, { "grad_norm": 0.13694755733013153, "learning_rate": 4.099012557657882e-05, "loss": 0.0118, "step": 57980 }, { "grad_norm": 0.4094238877296448, "learning_rate": 4.09738620368951e-05, "loss": 0.0156, "step": 57990 }, { "grad_norm": 0.1580846607685089, "learning_rate": 4.095759948429626e-05, "loss": 0.0107, "step": 58000 }, { "grad_norm": 0.09103621542453766, "learning_rate": 4.094133792056075e-05, "loss": 0.0136, "step": 58010 }, { "grad_norm": 0.13207148015499115, "learning_rate": 4.092507734746688e-05, "loss": 0.0164, "step": 58020 }, { "grad_norm": 0.22729024291038513, "learning_rate": 4.090881776679293e-05, "loss": 0.0147, "step": 58030 }, { "grad_norm": 0.26925361156463623, "learning_rate": 4.0892559180316994e-05, "loss": 0.0153, "step": 58040 }, { "grad_norm": 0.1701619178056717, "learning_rate": 4.08763015898171e-05, "loss": 0.0169, "step": 58050 }, { "grad_norm": 0.28150466084480286, "learning_rate": 4.0860044997071145e-05, "loss": 0.0214, "step": 58060 }, { "grad_norm": 0.45686817169189453, "learning_rate": 4.084378940385693e-05, "loss": 0.0169, "step": 58070 }, { "grad_norm": 0.11143657565116882, "learning_rate": 4.082753481195211e-05, "loss": 0.0139, "step": 58080 }, { "grad_norm": 0.12068832665681839, "learning_rate": 4.0811281223134324e-05, "loss": 0.0136, "step": 58090 }, { "grad_norm": 0.12653551995754242, "learning_rate": 4.0795028639181e-05, "loss": 0.0212, "step": 58100 }, { "grad_norm": 0.1751146763563156, "learning_rate": 4.077877706186951e-05, "loss": 0.0125, "step": 58110 }, { "grad_norm": 0.4217640161514282, "learning_rate": 4.076252649297709e-05, "loss": 0.0211, "step": 58120 }, { "grad_norm": 0.21910110116004944, "learning_rate": 4.07462769342809e-05, "loss": 0.0135, "step": 58130 }, { "grad_norm": 0.1874958723783493, "learning_rate": 4.073002838755793e-05, "loss": 0.0168, "step": 58140 }, { "grad_norm": 0.166451558470726, "learning_rate": 4.0713780854585115e-05, "loss": 0.0141, "step": 58150 }, { "grad_norm": 0.22456932067871094, "learning_rate": 4.069753433713927e-05, "loss": 0.0156, "step": 58160 }, { "grad_norm": 0.2810690999031067, "learning_rate": 4.068128883699709e-05, "loss": 0.0132, "step": 58170 }, { "grad_norm": 0.11232218146324158, "learning_rate": 4.0665044355935155e-05, "loss": 0.0155, "step": 58180 }, { "grad_norm": 0.3221534490585327, "learning_rate": 4.064880089572993e-05, "loss": 0.0178, "step": 58190 }, { "grad_norm": 0.27704137563705444, "learning_rate": 4.063255845815779e-05, "loss": 0.03, "step": 58200 }, { "grad_norm": 0.2833244204521179, "learning_rate": 4.0616317044994945e-05, "loss": 0.0178, "step": 58210 }, { "grad_norm": 0.21197423338890076, "learning_rate": 4.0600076658017585e-05, "loss": 0.0193, "step": 58220 }, { "grad_norm": 0.2895045280456543, "learning_rate": 4.0583837299001706e-05, "loss": 0.0189, "step": 58230 }, { "grad_norm": 0.12986181676387787, "learning_rate": 4.056759896972322e-05, "loss": 0.0175, "step": 58240 }, { "grad_norm": 0.180621936917305, "learning_rate": 4.055136167195793e-05, "loss": 0.0187, "step": 58250 }, { "grad_norm": 0.32781603932380676, "learning_rate": 4.053512540748152e-05, "loss": 0.0215, "step": 58260 }, { "grad_norm": 0.09050365537405014, "learning_rate": 4.051889017806955e-05, "loss": 0.0159, "step": 58270 }, { "grad_norm": 0.24535509943962097, "learning_rate": 4.0502655985497516e-05, "loss": 0.0165, "step": 58280 }, { "grad_norm": 0.20664697885513306, "learning_rate": 4.048642283154074e-05, "loss": 0.0172, "step": 58290 }, { "grad_norm": 0.23354259133338928, "learning_rate": 4.0470190717974464e-05, "loss": 0.0189, "step": 58300 }, { "grad_norm": 0.244118794798851, "learning_rate": 4.045395964657381e-05, "loss": 0.0247, "step": 58310 }, { "grad_norm": 0.1461118757724762, "learning_rate": 4.043772961911376e-05, "loss": 0.0145, "step": 58320 }, { "grad_norm": 0.3545408546924591, "learning_rate": 4.042150063736924e-05, "loss": 0.023, "step": 58330 }, { "grad_norm": 0.42198485136032104, "learning_rate": 4.0405272703114976e-05, "loss": 0.0131, "step": 58340 }, { "grad_norm": 0.2189338058233261, "learning_rate": 4.038904581812569e-05, "loss": 0.0122, "step": 58350 }, { "grad_norm": 0.10179868340492249, "learning_rate": 4.03728199841759e-05, "loss": 0.0136, "step": 58360 }, { "grad_norm": 0.364323228597641, "learning_rate": 4.0356595203040044e-05, "loss": 0.0175, "step": 58370 }, { "grad_norm": 0.12100128829479218, "learning_rate": 4.0340371476492426e-05, "loss": 0.0144, "step": 58380 }, { "grad_norm": 0.323801726102829, "learning_rate": 4.032414880630727e-05, "loss": 0.0252, "step": 58390 }, { "grad_norm": 0.1272614300251007, "learning_rate": 4.0307927194258646e-05, "loss": 0.0143, "step": 58400 }, { "grad_norm": 0.2142333984375, "learning_rate": 4.029170664212054e-05, "loss": 0.0138, "step": 58410 }, { "grad_norm": 0.13104493916034698, "learning_rate": 4.0275487151666794e-05, "loss": 0.0182, "step": 58420 }, { "grad_norm": 0.21427083015441895, "learning_rate": 4.025926872467116e-05, "loss": 0.0254, "step": 58430 }, { "grad_norm": 0.32969000935554504, "learning_rate": 4.024305136290725e-05, "loss": 0.0177, "step": 58440 }, { "grad_norm": 0.12982749938964844, "learning_rate": 4.022683506814858e-05, "loss": 0.0154, "step": 58450 }, { "grad_norm": 0.2008037567138672, "learning_rate": 4.021061984216853e-05, "loss": 0.0126, "step": 58460 }, { "grad_norm": 0.17376428842544556, "learning_rate": 4.019440568674035e-05, "loss": 0.0139, "step": 58470 }, { "grad_norm": 0.19605140388011932, "learning_rate": 4.017819260363724e-05, "loss": 0.0148, "step": 58480 }, { "grad_norm": 0.1360890120267868, "learning_rate": 4.016198059463223e-05, "loss": 0.0274, "step": 58490 }, { "grad_norm": 0.22252824902534485, "learning_rate": 4.0145769661498233e-05, "loss": 0.0211, "step": 58500 }, { "grad_norm": 0.307837575674057, "learning_rate": 4.012955980600804e-05, "loss": 0.0155, "step": 58510 }, { "grad_norm": 0.2152141034603119, "learning_rate": 4.0113351029934355e-05, "loss": 0.0164, "step": 58520 }, { "grad_norm": 0.22553302347660065, "learning_rate": 4.0097143335049714e-05, "loss": 0.0236, "step": 58530 }, { "grad_norm": 0.30563676357269287, "learning_rate": 4.0080936723126606e-05, "loss": 0.021, "step": 58540 }, { "grad_norm": 0.1371876299381256, "learning_rate": 4.006473119593735e-05, "loss": 0.0174, "step": 58550 }, { "grad_norm": 0.1946474015712738, "learning_rate": 4.004852675525415e-05, "loss": 0.0088, "step": 58560 }, { "grad_norm": 0.20096264779567719, "learning_rate": 4.00323234028491e-05, "loss": 0.014, "step": 58570 }, { "grad_norm": 0.1687348335981369, "learning_rate": 4.0016121140494186e-05, "loss": 0.0202, "step": 58580 }, { "grad_norm": 0.08777718245983124, "learning_rate": 3.999991996996122e-05, "loss": 0.0137, "step": 58590 }, { "grad_norm": 0.20332267880439758, "learning_rate": 3.9983719893022e-05, "loss": 0.0096, "step": 58600 }, { "grad_norm": 0.16970013082027435, "learning_rate": 3.996752091144812e-05, "loss": 0.0131, "step": 58610 }, { "grad_norm": 0.4345191419124603, "learning_rate": 3.9951323027011055e-05, "loss": 0.0186, "step": 58620 }, { "grad_norm": 0.23778128623962402, "learning_rate": 3.993512624148219e-05, "loss": 0.0213, "step": 58630 }, { "grad_norm": 0.21090149879455566, "learning_rate": 3.99189305566328e-05, "loss": 0.0147, "step": 58640 }, { "grad_norm": 0.09864111989736557, "learning_rate": 3.990273597423401e-05, "loss": 0.0136, "step": 58650 }, { "grad_norm": 0.15860067307949066, "learning_rate": 3.988654249605681e-05, "loss": 0.0098, "step": 58660 }, { "grad_norm": 0.12296555936336517, "learning_rate": 3.987035012387213e-05, "loss": 0.0192, "step": 58670 }, { "grad_norm": 0.4800270199775696, "learning_rate": 3.9854158859450735e-05, "loss": 0.0119, "step": 58680 }, { "grad_norm": 0.26609206199645996, "learning_rate": 3.983796870456328e-05, "loss": 0.024, "step": 58690 }, { "grad_norm": 0.23998995125293732, "learning_rate": 3.982177966098028e-05, "loss": 0.0191, "step": 58700 }, { "grad_norm": 0.09741048514842987, "learning_rate": 3.980559173047216e-05, "loss": 0.0216, "step": 58710 }, { "grad_norm": 0.23405541479587555, "learning_rate": 3.9789404914809176e-05, "loss": 0.0155, "step": 58720 }, { "grad_norm": 0.15183478593826294, "learning_rate": 3.9773219215761544e-05, "loss": 0.0109, "step": 58730 }, { "grad_norm": 0.15403124690055847, "learning_rate": 3.975703463509928e-05, "loss": 0.014, "step": 58740 }, { "grad_norm": 0.3530872166156769, "learning_rate": 3.974085117459231e-05, "loss": 0.0201, "step": 58750 }, { "grad_norm": 0.386917382478714, "learning_rate": 3.972466883601043e-05, "loss": 0.0173, "step": 58760 }, { "grad_norm": 0.23617549240589142, "learning_rate": 3.9708487621123316e-05, "loss": 0.0229, "step": 58770 }, { "grad_norm": 0.2170826643705368, "learning_rate": 3.969230753170051e-05, "loss": 0.0225, "step": 58780 }, { "grad_norm": 0.23765911161899567, "learning_rate": 3.967612856951146e-05, "loss": 0.0144, "step": 58790 }, { "grad_norm": 0.19666734337806702, "learning_rate": 3.9659950736325476e-05, "loss": 0.0131, "step": 58800 }, { "grad_norm": 0.28549250960350037, "learning_rate": 3.964377403391174e-05, "loss": 0.0376, "step": 58810 }, { "grad_norm": 0.12959106266498566, "learning_rate": 3.9627598464039303e-05, "loss": 0.016, "step": 58820 }, { "grad_norm": 0.12378012388944626, "learning_rate": 3.9611424028477096e-05, "loss": 0.0192, "step": 58830 }, { "grad_norm": 0.1371862143278122, "learning_rate": 3.959525072899394e-05, "loss": 0.0218, "step": 58840 }, { "grad_norm": 0.2590506374835968, "learning_rate": 3.9579078567358506e-05, "loss": 0.0225, "step": 58850 }, { "grad_norm": 0.2096211165189743, "learning_rate": 3.956290754533939e-05, "loss": 0.0158, "step": 58860 }, { "grad_norm": 0.17012521624565125, "learning_rate": 3.954673766470501e-05, "loss": 0.0184, "step": 58870 }, { "grad_norm": 0.1678204983472824, "learning_rate": 3.9530568927223674e-05, "loss": 0.0179, "step": 58880 }, { "grad_norm": 0.1400243192911148, "learning_rate": 3.95144013346636e-05, "loss": 0.0112, "step": 58890 }, { "grad_norm": 0.1842932105064392, "learning_rate": 3.949823488879282e-05, "loss": 0.0126, "step": 58900 }, { "grad_norm": 0.1651979386806488, "learning_rate": 3.9482069591379276e-05, "loss": 0.0225, "step": 58910 }, { "grad_norm": 0.18806365132331848, "learning_rate": 3.9465905444190784e-05, "loss": 0.0231, "step": 58920 }, { "grad_norm": 0.13586251437664032, "learning_rate": 3.944974244899505e-05, "loss": 0.0252, "step": 58930 }, { "grad_norm": 0.21477918326854706, "learning_rate": 3.9433580607559616e-05, "loss": 0.019, "step": 58940 }, { "grad_norm": 0.1808890402317047, "learning_rate": 3.941741992165193e-05, "loss": 0.0303, "step": 58950 }, { "grad_norm": 0.21684907376766205, "learning_rate": 3.940126039303928e-05, "loss": 0.01, "step": 58960 }, { "grad_norm": 0.17060637474060059, "learning_rate": 3.938510202348886e-05, "loss": 0.0236, "step": 58970 }, { "grad_norm": 0.20082329213619232, "learning_rate": 3.93689448147677e-05, "loss": 0.0146, "step": 58980 }, { "grad_norm": 0.40215378999710083, "learning_rate": 3.935278876864278e-05, "loss": 0.0186, "step": 58990 }, { "grad_norm": 0.13881182670593262, "learning_rate": 3.933663388688087e-05, "loss": 0.0258, "step": 59000 }, { "grad_norm": 0.2883225679397583, "learning_rate": 3.9320480171248645e-05, "loss": 0.0129, "step": 59010 }, { "grad_norm": 0.24006909132003784, "learning_rate": 3.930432762351265e-05, "loss": 0.0281, "step": 59020 }, { "grad_norm": 0.21806752681732178, "learning_rate": 3.928817624543931e-05, "loss": 0.0207, "step": 59030 }, { "grad_norm": 0.3946723937988281, "learning_rate": 3.9272026038794885e-05, "loss": 0.0212, "step": 59040 }, { "grad_norm": 0.23494401574134827, "learning_rate": 3.9255877005345586e-05, "loss": 0.0244, "step": 59050 }, { "grad_norm": 0.11180516332387924, "learning_rate": 3.923972914685743e-05, "loss": 0.0165, "step": 59060 }, { "grad_norm": 0.16179592907428741, "learning_rate": 3.922358246509631e-05, "loss": 0.0293, "step": 59070 }, { "grad_norm": 0.16327589750289917, "learning_rate": 3.920743696182801e-05, "loss": 0.0149, "step": 59080 }, { "grad_norm": 0.1334979236125946, "learning_rate": 3.9191292638818164e-05, "loss": 0.0217, "step": 59090 }, { "grad_norm": 0.2901095449924469, "learning_rate": 3.9175149497832316e-05, "loss": 0.0197, "step": 59100 }, { "grad_norm": 0.37324458360671997, "learning_rate": 3.915900754063584e-05, "loss": 0.0174, "step": 59110 }, { "grad_norm": 0.1552545726299286, "learning_rate": 3.9142866768994014e-05, "loss": 0.0135, "step": 59120 }, { "grad_norm": 0.18675397336483002, "learning_rate": 3.912672718467193e-05, "loss": 0.0155, "step": 59130 }, { "grad_norm": 0.33495083451271057, "learning_rate": 3.911058878943463e-05, "loss": 0.011, "step": 59140 }, { "grad_norm": 0.08329030126333237, "learning_rate": 3.9094451585046974e-05, "loss": 0.0164, "step": 59150 }, { "grad_norm": 0.1855757236480713, "learning_rate": 3.90783155732737e-05, "loss": 0.0172, "step": 59160 }, { "grad_norm": 0.21670743823051453, "learning_rate": 3.9062180755879373e-05, "loss": 0.013, "step": 59170 }, { "grad_norm": 0.2922905683517456, "learning_rate": 3.904604713462855e-05, "loss": 0.0087, "step": 59180 }, { "grad_norm": 0.08905592560768127, "learning_rate": 3.902991471128554e-05, "loss": 0.0145, "step": 59190 }, { "grad_norm": 0.06421717256307602, "learning_rate": 3.901378348761456e-05, "loss": 0.0145, "step": 59200 }, { "grad_norm": 0.27442124485969543, "learning_rate": 3.899765346537971e-05, "loss": 0.0134, "step": 59210 }, { "grad_norm": 0.1903281956911087, "learning_rate": 3.898152464634493e-05, "loss": 0.0174, "step": 59220 }, { "grad_norm": 0.08647805452346802, "learning_rate": 3.896539703227402e-05, "loss": 0.0167, "step": 59230 }, { "grad_norm": 0.5689562559127808, "learning_rate": 3.894927062493073e-05, "loss": 0.0167, "step": 59240 }, { "grad_norm": 0.16795262694358826, "learning_rate": 3.893314542607859e-05, "loss": 0.0199, "step": 59250 }, { "grad_norm": 0.1843058317899704, "learning_rate": 3.891702143748103e-05, "loss": 0.0157, "step": 59260 }, { "grad_norm": 0.08969958126544952, "learning_rate": 3.890089866090133e-05, "loss": 0.0134, "step": 59270 }, { "grad_norm": 0.25720837712287903, "learning_rate": 3.888477709810267e-05, "loss": 0.0194, "step": 59280 }, { "grad_norm": 0.1634104698896408, "learning_rate": 3.886865675084806e-05, "loss": 0.0181, "step": 59290 }, { "grad_norm": 0.19718514382839203, "learning_rate": 3.885253762090043e-05, "loss": 0.0169, "step": 59300 }, { "grad_norm": 0.13614177703857422, "learning_rate": 3.8836419710022534e-05, "loss": 0.016, "step": 59310 }, { "grad_norm": 0.2447417974472046, "learning_rate": 3.882030301997698e-05, "loss": 0.0134, "step": 59320 }, { "grad_norm": 0.27967363595962524, "learning_rate": 3.880418755252629e-05, "loss": 0.0158, "step": 59330 }, { "grad_norm": 0.17039954662322998, "learning_rate": 3.87880733094328e-05, "loss": 0.0109, "step": 59340 }, { "grad_norm": 0.1263718605041504, "learning_rate": 3.877196029245877e-05, "loss": 0.0104, "step": 59350 }, { "grad_norm": 0.10950029641389847, "learning_rate": 3.875584850336627e-05, "loss": 0.0169, "step": 59360 }, { "grad_norm": 0.2881889045238495, "learning_rate": 3.873973794391728e-05, "loss": 0.0132, "step": 59370 }, { "grad_norm": 0.2039555460214615, "learning_rate": 3.872362861587361e-05, "loss": 0.0147, "step": 59380 }, { "grad_norm": 0.29704245924949646, "learning_rate": 3.870752052099698e-05, "loss": 0.0088, "step": 59390 }, { "grad_norm": 0.119052954018116, "learning_rate": 3.8691413661048916e-05, "loss": 0.0105, "step": 59400 }, { "grad_norm": 0.8114023208618164, "learning_rate": 3.867530803779085e-05, "loss": 0.0156, "step": 59410 }, { "grad_norm": 0.2157435566186905, "learning_rate": 3.8659203652984054e-05, "loss": 0.0201, "step": 59420 }, { "grad_norm": 0.24065838754177094, "learning_rate": 3.864310050838972e-05, "loss": 0.0176, "step": 59430 }, { "grad_norm": 0.2556472718715668, "learning_rate": 3.8626998605768835e-05, "loss": 0.0147, "step": 59440 }, { "grad_norm": 0.2720741629600525, "learning_rate": 3.8610897946882294e-05, "loss": 0.0168, "step": 59450 }, { "grad_norm": 0.24382764101028442, "learning_rate": 3.859479853349082e-05, "loss": 0.0139, "step": 59460 }, { "grad_norm": 0.1838235706090927, "learning_rate": 3.857870036735503e-05, "loss": 0.0144, "step": 59470 }, { "grad_norm": 0.15736737847328186, "learning_rate": 3.856260345023538e-05, "loss": 0.0202, "step": 59480 }, { "grad_norm": 0.2172103375196457, "learning_rate": 3.854650778389224e-05, "loss": 0.0143, "step": 59490 }, { "grad_norm": 0.13514846563339233, "learning_rate": 3.8530413370085786e-05, "loss": 0.0172, "step": 59500 }, { "grad_norm": 0.30359703302383423, "learning_rate": 3.851432021057608e-05, "loss": 0.0193, "step": 59510 }, { "grad_norm": 0.19952116906642914, "learning_rate": 3.849822830712306e-05, "loss": 0.0187, "step": 59520 }, { "grad_norm": 0.3555344045162201, "learning_rate": 3.848213766148649e-05, "loss": 0.0237, "step": 59530 }, { "grad_norm": 0.1473616659641266, "learning_rate": 3.846604827542602e-05, "loss": 0.0108, "step": 59540 }, { "grad_norm": 0.12896358966827393, "learning_rate": 3.844996015070117e-05, "loss": 0.0225, "step": 59550 }, { "grad_norm": 0.24921002984046936, "learning_rate": 3.843387328907132e-05, "loss": 0.0154, "step": 59560 }, { "grad_norm": 0.2642747759819031, "learning_rate": 3.8417787692295696e-05, "loss": 0.0232, "step": 59570 }, { "grad_norm": 0.11667589098215103, "learning_rate": 3.84017033621334e-05, "loss": 0.0134, "step": 59580 }, { "grad_norm": 0.15956245362758636, "learning_rate": 3.838562030034338e-05, "loss": 0.0085, "step": 59590 }, { "grad_norm": 0.11218942701816559, "learning_rate": 3.836953850868447e-05, "loss": 0.013, "step": 59600 }, { "grad_norm": 0.2566230595111847, "learning_rate": 3.835345798891533e-05, "loss": 0.0138, "step": 59610 }, { "grad_norm": 0.3159613311290741, "learning_rate": 3.833737874279452e-05, "loss": 0.0173, "step": 59620 }, { "grad_norm": 0.22638647258281708, "learning_rate": 3.8321300772080433e-05, "loss": 0.0144, "step": 59630 }, { "grad_norm": 0.09192711859941483, "learning_rate": 3.8305224078531345e-05, "loss": 0.0142, "step": 59640 }, { "grad_norm": 0.24100252985954285, "learning_rate": 3.828914866390537e-05, "loss": 0.0208, "step": 59650 }, { "grad_norm": 0.459335595369339, "learning_rate": 3.827307452996048e-05, "loss": 0.03, "step": 59660 }, { "grad_norm": 0.11543852835893631, "learning_rate": 3.825700167845454e-05, "loss": 0.0196, "step": 59670 }, { "grad_norm": 0.11628147214651108, "learning_rate": 3.8240930111145214e-05, "loss": 0.016, "step": 59680 }, { "grad_norm": 0.11256903409957886, "learning_rate": 3.8224859829790116e-05, "loss": 0.012, "step": 59690 }, { "grad_norm": 0.1418120265007019, "learning_rate": 3.820879083614664e-05, "loss": 0.0197, "step": 59700 }, { "grad_norm": 0.13361220061779022, "learning_rate": 3.819272313197208e-05, "loss": 0.0191, "step": 59710 }, { "grad_norm": 0.21230746805667877, "learning_rate": 3.817665671902355e-05, "loss": 0.0117, "step": 59720 }, { "grad_norm": 0.08524931967258453, "learning_rate": 3.816059159905808e-05, "loss": 0.0116, "step": 59730 }, { "grad_norm": 0.1342272013425827, "learning_rate": 3.8144527773832476e-05, "loss": 0.0217, "step": 59740 }, { "grad_norm": 0.5099042057991028, "learning_rate": 3.812846524510352e-05, "loss": 0.015, "step": 59750 }, { "grad_norm": 0.13174556195735931, "learning_rate": 3.811240401462775e-05, "loss": 0.0175, "step": 59760 }, { "grad_norm": 0.1147262454032898, "learning_rate": 3.809634408416162e-05, "loss": 0.0205, "step": 59770 }, { "grad_norm": 0.21675540506839752, "learning_rate": 3.8080285455461394e-05, "loss": 0.0147, "step": 59780 }, { "grad_norm": 0.16551634669303894, "learning_rate": 3.806422813028323e-05, "loss": 0.0156, "step": 59790 }, { "grad_norm": 0.29469290375709534, "learning_rate": 3.8048172110383114e-05, "loss": 0.0142, "step": 59800 }, { "grad_norm": 0.13956588506698608, "learning_rate": 3.8032117397516944e-05, "loss": 0.0097, "step": 59810 }, { "grad_norm": 0.31340262293815613, "learning_rate": 3.801606399344042e-05, "loss": 0.0125, "step": 59820 }, { "grad_norm": 0.20414578914642334, "learning_rate": 3.800001189990913e-05, "loss": 0.0121, "step": 59830 }, { "grad_norm": 0.24518759548664093, "learning_rate": 3.798396111867847e-05, "loss": 0.0148, "step": 59840 }, { "grad_norm": 0.3696480989456177, "learning_rate": 3.796791165150378e-05, "loss": 0.0154, "step": 59850 }, { "grad_norm": 0.3176289200782776, "learning_rate": 3.795186350014018e-05, "loss": 0.0148, "step": 59860 }, { "grad_norm": 0.21013472974300385, "learning_rate": 3.793581666634266e-05, "loss": 0.019, "step": 59870 }, { "grad_norm": 0.14587195217609406, "learning_rate": 3.791977115186609e-05, "loss": 0.0155, "step": 59880 }, { "grad_norm": 0.2221824824810028, "learning_rate": 3.79037269584652e-05, "loss": 0.016, "step": 59890 }, { "grad_norm": 0.17458084225654602, "learning_rate": 3.7887684087894545e-05, "loss": 0.016, "step": 59900 }, { "grad_norm": 0.2746345102787018, "learning_rate": 3.787164254190854e-05, "loss": 0.0187, "step": 59910 }, { "grad_norm": 0.2641327679157257, "learning_rate": 3.7855602322261473e-05, "loss": 0.0155, "step": 59920 }, { "grad_norm": 0.11156832426786423, "learning_rate": 3.783956343070746e-05, "loss": 0.0235, "step": 59930 }, { "grad_norm": 0.1670355498790741, "learning_rate": 3.7823525869000526e-05, "loss": 0.0219, "step": 59940 }, { "grad_norm": 0.15778861939907074, "learning_rate": 3.78074896388945e-05, "loss": 0.0211, "step": 59950 }, { "grad_norm": 0.15889310836791992, "learning_rate": 3.7791454742143075e-05, "loss": 0.0209, "step": 59960 }, { "grad_norm": 0.14771254360675812, "learning_rate": 3.7775421180499795e-05, "loss": 0.0148, "step": 59970 }, { "grad_norm": 0.17125912010669708, "learning_rate": 3.7759388955718086e-05, "loss": 0.0136, "step": 59980 }, { "grad_norm": 0.26356276869773865, "learning_rate": 3.774335806955117e-05, "loss": 0.0213, "step": 59990 }, { "grad_norm": 0.3009594976902008, "learning_rate": 3.7727328523752215e-05, "loss": 0.0177, "step": 60000 }, { "grad_norm": 0.19017356634140015, "learning_rate": 3.771130032007415e-05, "loss": 0.0114, "step": 60010 }, { "grad_norm": 0.23078300058841705, "learning_rate": 3.769527346026982e-05, "loss": 0.0135, "step": 60020 }, { "grad_norm": 0.16731858253479004, "learning_rate": 3.767924794609188e-05, "loss": 0.0129, "step": 60030 }, { "grad_norm": 0.07429671287536621, "learning_rate": 3.7663223779292854e-05, "loss": 0.0182, "step": 60040 }, { "grad_norm": 0.1642143428325653, "learning_rate": 3.764720096162512e-05, "loss": 0.0133, "step": 60050 }, { "grad_norm": 0.43217378854751587, "learning_rate": 3.763117949484092e-05, "loss": 0.0152, "step": 60060 }, { "grad_norm": 0.29548901319503784, "learning_rate": 3.7615159380692344e-05, "loss": 0.0162, "step": 60070 }, { "grad_norm": 0.10948202013969421, "learning_rate": 3.759914062093132e-05, "loss": 0.0118, "step": 60080 }, { "grad_norm": 0.22686423361301422, "learning_rate": 3.7583123217309615e-05, "loss": 0.0139, "step": 60090 }, { "grad_norm": 0.45901843905448914, "learning_rate": 3.7567107171578904e-05, "loss": 0.0259, "step": 60100 }, { "grad_norm": 0.26255109906196594, "learning_rate": 3.755109248549066e-05, "loss": 0.0146, "step": 60110 }, { "grad_norm": 0.19523847103118896, "learning_rate": 3.75350791607962e-05, "loss": 0.0109, "step": 60120 }, { "grad_norm": 0.16195710003376007, "learning_rate": 3.751906719924676e-05, "loss": 0.0114, "step": 60130 }, { "grad_norm": 0.12671442329883575, "learning_rate": 3.750305660259337e-05, "loss": 0.0196, "step": 60140 }, { "grad_norm": 0.12834285199642181, "learning_rate": 3.748704737258693e-05, "loss": 0.0165, "step": 60150 }, { "grad_norm": 0.27221226692199707, "learning_rate": 3.747103951097816e-05, "loss": 0.0225, "step": 60160 }, { "grad_norm": 0.2217487245798111, "learning_rate": 3.745503301951768e-05, "loss": 0.0171, "step": 60170 }, { "grad_norm": 0.2133711278438568, "learning_rate": 3.743902789995592e-05, "loss": 0.0179, "step": 60180 }, { "grad_norm": 0.3384716808795929, "learning_rate": 3.7423024154043183e-05, "loss": 0.0143, "step": 60190 }, { "grad_norm": 0.09824196249246597, "learning_rate": 3.740702178352963e-05, "loss": 0.0101, "step": 60200 }, { "grad_norm": 0.14972615242004395, "learning_rate": 3.739102079016523e-05, "loss": 0.0145, "step": 60210 }, { "grad_norm": 0.1324114352464676, "learning_rate": 3.7375021175699846e-05, "loss": 0.0151, "step": 60220 }, { "grad_norm": 0.1465110331773758, "learning_rate": 3.735902294188317e-05, "loss": 0.0232, "step": 60230 }, { "grad_norm": 0.21606361865997314, "learning_rate": 3.734302609046473e-05, "loss": 0.0145, "step": 60240 }, { "grad_norm": 0.4241209924221039, "learning_rate": 3.732703062319391e-05, "loss": 0.0215, "step": 60250 }, { "grad_norm": 0.14236465096473694, "learning_rate": 3.731103654181999e-05, "loss": 0.0169, "step": 60260 }, { "grad_norm": 0.16202044486999512, "learning_rate": 3.729504384809203e-05, "loss": 0.018, "step": 60270 }, { "grad_norm": 0.25600486993789673, "learning_rate": 3.727905254375898e-05, "loss": 0.0136, "step": 60280 }, { "grad_norm": 0.28810739517211914, "learning_rate": 3.7263062630569614e-05, "loss": 0.0176, "step": 60290 }, { "grad_norm": 0.16907916963100433, "learning_rate": 3.724707411027256e-05, "loss": 0.016, "step": 60300 }, { "grad_norm": 0.3269191384315491, "learning_rate": 3.723108698461631e-05, "loss": 0.0175, "step": 60310 }, { "grad_norm": 0.08413272351026535, "learning_rate": 3.72151012553492e-05, "loss": 0.0142, "step": 60320 }, { "grad_norm": 0.248578280210495, "learning_rate": 3.719911692421939e-05, "loss": 0.0128, "step": 60330 }, { "grad_norm": 0.11716748774051666, "learning_rate": 3.71831339929749e-05, "loss": 0.0148, "step": 60340 }, { "grad_norm": 0.10401267558336258, "learning_rate": 3.716715246336362e-05, "loss": 0.0133, "step": 60350 }, { "grad_norm": 0.2603703737258911, "learning_rate": 3.715117233713324e-05, "loss": 0.0183, "step": 60360 }, { "grad_norm": 0.1959637999534607, "learning_rate": 3.713519361603135e-05, "loss": 0.0122, "step": 60370 }, { "grad_norm": 0.2876969277858734, "learning_rate": 3.711921630180532e-05, "loss": 0.016, "step": 60380 }, { "grad_norm": 0.1672433614730835, "learning_rate": 3.710324039620245e-05, "loss": 0.0164, "step": 60390 }, { "grad_norm": 0.2522776126861572, "learning_rate": 3.708726590096982e-05, "loss": 0.0151, "step": 60400 }, { "grad_norm": 0.09324757754802704, "learning_rate": 3.707129281785437e-05, "loss": 0.0116, "step": 60410 }, { "grad_norm": 0.4798603355884552, "learning_rate": 3.705532114860291e-05, "loss": 0.0249, "step": 60420 }, { "grad_norm": 0.25130876898765564, "learning_rate": 3.703935089496207e-05, "loss": 0.0249, "step": 60430 }, { "grad_norm": 0.2952900528907776, "learning_rate": 3.702338205867829e-05, "loss": 0.0234, "step": 60440 }, { "grad_norm": 0.3031642436981201, "learning_rate": 3.7007414641497976e-05, "loss": 0.0198, "step": 60450 }, { "grad_norm": 0.40933001041412354, "learning_rate": 3.699144864516726e-05, "loss": 0.028, "step": 60460 }, { "grad_norm": 0.156361386179924, "learning_rate": 3.6975484071432153e-05, "loss": 0.0107, "step": 60470 }, { "grad_norm": 0.27293524146080017, "learning_rate": 3.695952092203854e-05, "loss": 0.0115, "step": 60480 }, { "grad_norm": 0.29516011476516724, "learning_rate": 3.69435591987321e-05, "loss": 0.0221, "step": 60490 }, { "grad_norm": 0.2829095125198364, "learning_rate": 3.6927598903258374e-05, "loss": 0.0175, "step": 60500 }, { "grad_norm": 0.22342254221439362, "learning_rate": 3.691164003736281e-05, "loss": 0.0145, "step": 60510 }, { "grad_norm": 0.218134805560112, "learning_rate": 3.689568260279061e-05, "loss": 0.0166, "step": 60520 }, { "grad_norm": 0.22752948105335236, "learning_rate": 3.687972660128686e-05, "loss": 0.0079, "step": 60530 }, { "grad_norm": 0.2999168634414673, "learning_rate": 3.686377203459648e-05, "loss": 0.0196, "step": 60540 }, { "grad_norm": 0.1471831500530243, "learning_rate": 3.6847818904464226e-05, "loss": 0.0139, "step": 60550 }, { "grad_norm": 0.19644014537334442, "learning_rate": 3.683186721263474e-05, "loss": 0.0146, "step": 60560 }, { "grad_norm": 0.18234288692474365, "learning_rate": 3.6815916960852435e-05, "loss": 0.0179, "step": 60570 }, { "grad_norm": 0.12692207098007202, "learning_rate": 3.679996815086165e-05, "loss": 0.0174, "step": 60580 }, { "grad_norm": 0.20819006860256195, "learning_rate": 3.678402078440649e-05, "loss": 0.0155, "step": 60590 }, { "grad_norm": 0.20360350608825684, "learning_rate": 3.676807486323096e-05, "loss": 0.0098, "step": 60600 }, { "grad_norm": 0.16379371285438538, "learning_rate": 3.6752130389078864e-05, "loss": 0.0147, "step": 60610 }, { "grad_norm": 0.14699684083461761, "learning_rate": 3.673618736369388e-05, "loss": 0.0188, "step": 60620 }, { "grad_norm": 0.23167873919010162, "learning_rate": 3.6720245788819474e-05, "loss": 0.0141, "step": 60630 }, { "grad_norm": 0.18665848672389984, "learning_rate": 3.6704305666199044e-05, "loss": 0.0134, "step": 60640 }, { "grad_norm": 0.20543162524700165, "learning_rate": 3.668836699757576e-05, "loss": 0.0159, "step": 60650 }, { "grad_norm": 0.3086744546890259, "learning_rate": 3.667242978469265e-05, "loss": 0.0153, "step": 60660 }, { "grad_norm": 0.13745218515396118, "learning_rate": 3.6656494029292575e-05, "loss": 0.0161, "step": 60670 }, { "grad_norm": 0.09652228653430939, "learning_rate": 3.664055973311825e-05, "loss": 0.0173, "step": 60680 }, { "grad_norm": 0.11429198831319809, "learning_rate": 3.662462689791221e-05, "loss": 0.0181, "step": 60690 }, { "grad_norm": 0.2965574264526367, "learning_rate": 3.660869552541689e-05, "loss": 0.0163, "step": 60700 }, { "grad_norm": 0.1401084065437317, "learning_rate": 3.659276561737448e-05, "loss": 0.0182, "step": 60710 }, { "grad_norm": 0.13119542598724365, "learning_rate": 3.657683717552708e-05, "loss": 0.0167, "step": 60720 }, { "grad_norm": 0.12677398324012756, "learning_rate": 3.6560910201616574e-05, "loss": 0.0152, "step": 60730 }, { "grad_norm": 0.09764154255390167, "learning_rate": 3.654498469738472e-05, "loss": 0.0152, "step": 60740 }, { "grad_norm": 0.1255211979150772, "learning_rate": 3.6529060664573126e-05, "loss": 0.0111, "step": 60750 }, { "grad_norm": 0.09490232914686203, "learning_rate": 3.6513138104923176e-05, "loss": 0.0161, "step": 60760 }, { "grad_norm": 0.4552716910839081, "learning_rate": 3.6497217020176176e-05, "loss": 0.0139, "step": 60770 }, { "grad_norm": 0.33839893341064453, "learning_rate": 3.648129741207323e-05, "loss": 0.0224, "step": 60780 }, { "grad_norm": 0.14384956657886505, "learning_rate": 3.646537928235527e-05, "loss": 0.0174, "step": 60790 }, { "grad_norm": 0.23832552134990692, "learning_rate": 3.6449462632763075e-05, "loss": 0.018, "step": 60800 }, { "grad_norm": 0.17309588193893433, "learning_rate": 3.643354746503729e-05, "loss": 0.0144, "step": 60810 }, { "grad_norm": 0.15823695063591003, "learning_rate": 3.6417633780918345e-05, "loss": 0.0094, "step": 60820 }, { "grad_norm": 0.06653419137001038, "learning_rate": 3.640172158214655e-05, "loss": 0.0109, "step": 60830 }, { "grad_norm": 0.16479407250881195, "learning_rate": 3.6385810870462035e-05, "loss": 0.0171, "step": 60840 }, { "grad_norm": 0.17097419500350952, "learning_rate": 3.636990164760479e-05, "loss": 0.0168, "step": 60850 }, { "grad_norm": 0.13998839259147644, "learning_rate": 3.6353993915314596e-05, "loss": 0.0113, "step": 60860 }, { "grad_norm": 0.0847495049238205, "learning_rate": 3.6338087675331124e-05, "loss": 0.0121, "step": 60870 }, { "grad_norm": 0.29825302958488464, "learning_rate": 3.632218292939383e-05, "loss": 0.0178, "step": 60880 }, { "grad_norm": 0.2532911002635956, "learning_rate": 3.6306279679242033e-05, "loss": 0.0153, "step": 60890 }, { "grad_norm": 0.21051247417926788, "learning_rate": 3.629037792661492e-05, "loss": 0.0147, "step": 60900 }, { "grad_norm": 0.1929876059293747, "learning_rate": 3.627447767325147e-05, "loss": 0.0161, "step": 60910 }, { "grad_norm": 0.19794413447380066, "learning_rate": 3.62585789208905e-05, "loss": 0.0124, "step": 60920 }, { "grad_norm": 0.6326829791069031, "learning_rate": 3.624268167127068e-05, "loss": 0.0246, "step": 60930 }, { "grad_norm": 0.32470375299453735, "learning_rate": 3.622678592613051e-05, "loss": 0.0134, "step": 60940 }, { "grad_norm": 0.15060032904148102, "learning_rate": 3.6210891687208304e-05, "loss": 0.0178, "step": 60950 }, { "grad_norm": 0.07136324048042297, "learning_rate": 3.619499895624228e-05, "loss": 0.0122, "step": 60960 }, { "grad_norm": 0.16801272332668304, "learning_rate": 3.61791077349704e-05, "loss": 0.0219, "step": 60970 }, { "grad_norm": 0.19052919745445251, "learning_rate": 3.616321802513053e-05, "loss": 0.0275, "step": 60980 }, { "grad_norm": 0.27040955424308777, "learning_rate": 3.6147329828460325e-05, "loss": 0.0135, "step": 60990 }, { "grad_norm": 0.2686440348625183, "learning_rate": 3.6131443146697306e-05, "loss": 0.0171, "step": 61000 }, { "grad_norm": 0.19976316392421722, "learning_rate": 3.611555798157879e-05, "loss": 0.0158, "step": 61010 }, { "grad_norm": 0.16029340028762817, "learning_rate": 3.6099674334842e-05, "loss": 0.0147, "step": 61020 }, { "grad_norm": 0.2721351385116577, "learning_rate": 3.6083792208223934e-05, "loss": 0.0199, "step": 61030 }, { "grad_norm": 0.3198961615562439, "learning_rate": 3.606791160346142e-05, "loss": 0.0152, "step": 61040 }, { "grad_norm": 0.20981229841709137, "learning_rate": 3.6052032522291134e-05, "loss": 0.0169, "step": 61050 }, { "grad_norm": 0.3068089485168457, "learning_rate": 3.603615496644962e-05, "loss": 0.0204, "step": 61060 }, { "grad_norm": 0.35146182775497437, "learning_rate": 3.60202789376732e-05, "loss": 0.0202, "step": 61070 }, { "grad_norm": 0.1748577356338501, "learning_rate": 3.6004404437698036e-05, "loss": 0.0131, "step": 61080 }, { "grad_norm": 0.07784552127122879, "learning_rate": 3.5988531468260184e-05, "loss": 0.0137, "step": 61090 }, { "grad_norm": 0.17614951729774475, "learning_rate": 3.597266003109546e-05, "loss": 0.0144, "step": 61100 }, { "grad_norm": 0.10226471722126007, "learning_rate": 3.5956790127939557e-05, "loss": 0.0089, "step": 61110 }, { "grad_norm": 0.29250338673591614, "learning_rate": 3.594092176052796e-05, "loss": 0.0131, "step": 61120 }, { "grad_norm": 0.19781053066253662, "learning_rate": 3.592505493059603e-05, "loss": 0.0176, "step": 61130 }, { "grad_norm": 0.1814744621515274, "learning_rate": 3.5909189639878905e-05, "loss": 0.022, "step": 61140 }, { "grad_norm": 0.4923657476902008, "learning_rate": 3.5893325890111626e-05, "loss": 0.0189, "step": 61150 }, { "grad_norm": 0.16792990267276764, "learning_rate": 3.5877463683029034e-05, "loss": 0.0142, "step": 61160 }, { "grad_norm": 0.1612715721130371, "learning_rate": 3.586160302036578e-05, "loss": 0.0179, "step": 61170 }, { "grad_norm": 0.2490878701210022, "learning_rate": 3.584574390385636e-05, "loss": 0.0146, "step": 61180 }, { "grad_norm": 0.2807655930519104, "learning_rate": 3.582988633523511e-05, "loss": 0.0168, "step": 61190 }, { "grad_norm": 0.3541182577610016, "learning_rate": 3.581403031623616e-05, "loss": 0.0178, "step": 61200 }, { "grad_norm": 0.16471409797668457, "learning_rate": 3.579817584859354e-05, "loss": 0.0221, "step": 61210 }, { "grad_norm": 0.233845517039299, "learning_rate": 3.578232293404107e-05, "loss": 0.0177, "step": 61220 }, { "grad_norm": 0.16775557398796082, "learning_rate": 3.5766471574312385e-05, "loss": 0.0155, "step": 61230 }, { "grad_norm": 0.21070459485054016, "learning_rate": 3.5750621771140964e-05, "loss": 0.0291, "step": 61240 }, { "grad_norm": 0.27949464321136475, "learning_rate": 3.5734773526260115e-05, "loss": 0.0192, "step": 61250 }, { "grad_norm": 0.14714393019676208, "learning_rate": 3.5718926841402996e-05, "loss": 0.0169, "step": 61260 }, { "grad_norm": 0.26857805252075195, "learning_rate": 3.5703081718302555e-05, "loss": 0.0136, "step": 61270 }, { "grad_norm": 0.19234460592269897, "learning_rate": 3.568723815869162e-05, "loss": 0.0187, "step": 61280 }, { "grad_norm": 0.25449514389038086, "learning_rate": 3.5671396164302777e-05, "loss": 0.0263, "step": 61290 }, { "grad_norm": 0.3585469424724579, "learning_rate": 3.565555573686853e-05, "loss": 0.0232, "step": 61300 }, { "grad_norm": 0.20234152674674988, "learning_rate": 3.563971687812113e-05, "loss": 0.0234, "step": 61310 }, { "grad_norm": 0.3296876847743988, "learning_rate": 3.5623879589792706e-05, "loss": 0.0181, "step": 61320 }, { "grad_norm": 0.2011910229921341, "learning_rate": 3.5608043873615185e-05, "loss": 0.0166, "step": 61330 }, { "grad_norm": 0.1477389931678772, "learning_rate": 3.559220973132035e-05, "loss": 0.012, "step": 61340 }, { "grad_norm": 0.14064091444015503, "learning_rate": 3.557637716463981e-05, "loss": 0.0162, "step": 61350 }, { "grad_norm": 0.18749012053012848, "learning_rate": 3.556054617530497e-05, "loss": 0.0157, "step": 61360 }, { "grad_norm": 0.4467605650424957, "learning_rate": 3.5544716765047106e-05, "loss": 0.0236, "step": 61370 }, { "grad_norm": 0.29713889956474304, "learning_rate": 3.5528888935597274e-05, "loss": 0.019, "step": 61380 }, { "grad_norm": 0.12467499077320099, "learning_rate": 3.551306268868636e-05, "loss": 0.0125, "step": 61390 }, { "grad_norm": 0.12615567445755005, "learning_rate": 3.5497238026045166e-05, "loss": 0.0228, "step": 61400 }, { "grad_norm": 0.14531642198562622, "learning_rate": 3.548141494940421e-05, "loss": 0.013, "step": 61410 }, { "grad_norm": 0.2995226979255676, "learning_rate": 3.546559346049388e-05, "loss": 0.0176, "step": 61420 }, { "grad_norm": 0.1828271448612213, "learning_rate": 3.544977356104441e-05, "loss": 0.0148, "step": 61430 }, { "grad_norm": 0.20892967283725739, "learning_rate": 3.543395525278582e-05, "loss": 0.0135, "step": 61440 }, { "grad_norm": 0.23401641845703125, "learning_rate": 3.5418138537447985e-05, "loss": 0.016, "step": 61450 }, { "grad_norm": 0.08004608750343323, "learning_rate": 3.5402323416760574e-05, "loss": 0.0105, "step": 61460 }, { "grad_norm": 0.31038734316825867, "learning_rate": 3.538650989245316e-05, "loss": 0.0138, "step": 61470 }, { "grad_norm": 0.33995136618614197, "learning_rate": 3.537069796625504e-05, "loss": 0.0142, "step": 61480 }, { "grad_norm": 0.11295024305582047, "learning_rate": 3.53548876398954e-05, "loss": 0.0102, "step": 61490 }, { "grad_norm": 0.0494530014693737, "learning_rate": 3.5339078915103214e-05, "loss": 0.0094, "step": 61500 }, { "grad_norm": 0.28858375549316406, "learning_rate": 3.532327179360733e-05, "loss": 0.021, "step": 61510 }, { "grad_norm": 0.1635885387659073, "learning_rate": 3.530746627713636e-05, "loss": 0.0099, "step": 61520 }, { "grad_norm": 0.24662762880325317, "learning_rate": 3.5291662367418805e-05, "loss": 0.018, "step": 61530 }, { "grad_norm": 0.36120906472206116, "learning_rate": 3.527586006618293e-05, "loss": 0.0174, "step": 61540 }, { "grad_norm": 0.15211190283298492, "learning_rate": 3.526005937515687e-05, "loss": 0.0187, "step": 61550 }, { "grad_norm": 0.7500991225242615, "learning_rate": 3.524426029606856e-05, "loss": 0.0247, "step": 61560 }, { "grad_norm": 0.3176489472389221, "learning_rate": 3.5228462830645746e-05, "loss": 0.0123, "step": 61570 }, { "grad_norm": 0.3007149398326874, "learning_rate": 3.5212666980616036e-05, "loss": 0.0186, "step": 61580 }, { "grad_norm": 0.18629653751850128, "learning_rate": 3.519687274770681e-05, "loss": 0.0153, "step": 61590 }, { "grad_norm": 0.13978594541549683, "learning_rate": 3.518108013364535e-05, "loss": 0.0183, "step": 61600 }, { "grad_norm": 0.16582536697387695, "learning_rate": 3.516528914015868e-05, "loss": 0.0164, "step": 61610 }, { "grad_norm": 0.16448092460632324, "learning_rate": 3.5149499768973694e-05, "loss": 0.0176, "step": 61620 }, { "grad_norm": 0.2088346779346466, "learning_rate": 3.513371202181708e-05, "loss": 0.0076, "step": 61630 }, { "grad_norm": 0.4453124403953552, "learning_rate": 3.511792590041537e-05, "loss": 0.018, "step": 61640 }, { "grad_norm": 0.27611202001571655, "learning_rate": 3.510214140649488e-05, "loss": 0.0253, "step": 61650 }, { "grad_norm": 0.17157791554927826, "learning_rate": 3.508635854178184e-05, "loss": 0.0139, "step": 61660 }, { "grad_norm": 0.17015421390533447, "learning_rate": 3.50705773080022e-05, "loss": 0.0132, "step": 61670 }, { "grad_norm": 0.23803496360778809, "learning_rate": 3.505479770688178e-05, "loss": 0.0107, "step": 61680 }, { "grad_norm": 0.4286368787288666, "learning_rate": 3.503901974014622e-05, "loss": 0.0138, "step": 61690 }, { "grad_norm": 0.21503375470638275, "learning_rate": 3.502324340952095e-05, "loss": 0.0162, "step": 61700 }, { "grad_norm": 0.1213400661945343, "learning_rate": 3.500746871673126e-05, "loss": 0.0133, "step": 61710 }, { "grad_norm": 0.11753955483436584, "learning_rate": 3.499169566350227e-05, "loss": 0.0181, "step": 61720 }, { "grad_norm": 0.2328244149684906, "learning_rate": 3.4975924251558876e-05, "loss": 0.0118, "step": 61730 }, { "grad_norm": 0.08899760991334915, "learning_rate": 3.496015448262582e-05, "loss": 0.0202, "step": 61740 }, { "grad_norm": 0.1555224359035492, "learning_rate": 3.494438635842764e-05, "loss": 0.0124, "step": 61750 }, { "grad_norm": 0.23948940634727478, "learning_rate": 3.492861988068874e-05, "loss": 0.0157, "step": 61760 }, { "grad_norm": 0.1397279053926468, "learning_rate": 3.491285505113332e-05, "loss": 0.0143, "step": 61770 }, { "grad_norm": 0.2423011213541031, "learning_rate": 3.4897091871485376e-05, "loss": 0.0234, "step": 61780 }, { "grad_norm": 0.1359083503484726, "learning_rate": 3.488133034346875e-05, "loss": 0.0135, "step": 61790 }, { "grad_norm": 0.2040887176990509, "learning_rate": 3.486557046880713e-05, "loss": 0.0144, "step": 61800 }, { "grad_norm": 0.27375146746635437, "learning_rate": 3.484981224922395e-05, "loss": 0.014, "step": 61810 }, { "grad_norm": 0.1345285177230835, "learning_rate": 3.4834055686442544e-05, "loss": 0.013, "step": 61820 }, { "grad_norm": 0.34913432598114014, "learning_rate": 3.4818300782186e-05, "loss": 0.0155, "step": 61830 }, { "grad_norm": 0.12479540705680847, "learning_rate": 3.4802547538177236e-05, "loss": 0.0165, "step": 61840 }, { "grad_norm": 0.3633100390434265, "learning_rate": 3.478679595613905e-05, "loss": 0.016, "step": 61850 }, { "grad_norm": 0.1918424516916275, "learning_rate": 3.4771046037793975e-05, "loss": 0.0171, "step": 61860 }, { "grad_norm": 0.2124144732952118, "learning_rate": 3.4755297784864416e-05, "loss": 0.0147, "step": 61870 }, { "grad_norm": 0.25350305438041687, "learning_rate": 3.473955119907257e-05, "loss": 0.0117, "step": 61880 }, { "grad_norm": 0.3367510139942169, "learning_rate": 3.472380628214046e-05, "loss": 0.0185, "step": 61890 }, { "grad_norm": 0.2144034504890442, "learning_rate": 3.4708063035789906e-05, "loss": 0.0125, "step": 61900 }, { "grad_norm": 0.11390218883752823, "learning_rate": 3.4692321461742614e-05, "loss": 0.0097, "step": 61910 }, { "grad_norm": 0.22370144724845886, "learning_rate": 3.4676581561720023e-05, "loss": 0.0123, "step": 61920 }, { "grad_norm": 0.12218990921974182, "learning_rate": 3.466084333744344e-05, "loss": 0.0116, "step": 61930 }, { "grad_norm": 0.23023304343223572, "learning_rate": 3.4645106790633964e-05, "loss": 0.0164, "step": 61940 }, { "grad_norm": 0.17935527861118317, "learning_rate": 3.4629371923012526e-05, "loss": 0.0223, "step": 61950 }, { "grad_norm": 0.4074549674987793, "learning_rate": 3.4613638736299856e-05, "loss": 0.0186, "step": 61960 }, { "grad_norm": 0.20887164771556854, "learning_rate": 3.459790723221652e-05, "loss": 0.0126, "step": 61970 }, { "grad_norm": 0.30094850063323975, "learning_rate": 3.45821774124829e-05, "loss": 0.0169, "step": 61980 }, { "grad_norm": 0.24850846827030182, "learning_rate": 3.456644927881918e-05, "loss": 0.0166, "step": 61990 }, { "grad_norm": 0.2921043038368225, "learning_rate": 3.4550722832945354e-05, "loss": 0.02, "step": 62000 }, { "grad_norm": 0.20584334433078766, "learning_rate": 3.453499807658126e-05, "loss": 0.0122, "step": 62010 }, { "grad_norm": 0.13539735972881317, "learning_rate": 3.451927501144653e-05, "loss": 0.0158, "step": 62020 }, { "grad_norm": 0.4752326011657715, "learning_rate": 3.450355363926059e-05, "loss": 0.0109, "step": 62030 }, { "grad_norm": 0.08947013318538666, "learning_rate": 3.448783396174273e-05, "loss": 0.0176, "step": 62040 }, { "grad_norm": 0.4468851089477539, "learning_rate": 3.4472115980612034e-05, "loss": 0.0155, "step": 62050 }, { "grad_norm": 0.14841556549072266, "learning_rate": 3.4456399697587396e-05, "loss": 0.0133, "step": 62060 }, { "grad_norm": 0.1322304755449295, "learning_rate": 3.4440685114387506e-05, "loss": 0.0171, "step": 62070 }, { "grad_norm": 0.12777091562747955, "learning_rate": 3.442497223273091e-05, "loss": 0.0162, "step": 62080 }, { "grad_norm": 0.17554670572280884, "learning_rate": 3.4409261054335926e-05, "loss": 0.0238, "step": 62090 }, { "grad_norm": 0.15997827053070068, "learning_rate": 3.439355158092069e-05, "loss": 0.0288, "step": 62100 }, { "grad_norm": 0.25619247555732727, "learning_rate": 3.437784381420322e-05, "loss": 0.0147, "step": 62110 }, { "grad_norm": 0.15137341618537903, "learning_rate": 3.436213775590126e-05, "loss": 0.0126, "step": 62120 }, { "grad_norm": 0.3121117949485779, "learning_rate": 3.4346433407732405e-05, "loss": 0.022, "step": 62130 }, { "grad_norm": 0.24037663638591766, "learning_rate": 3.433073077141405e-05, "loss": 0.0165, "step": 62140 }, { "grad_norm": 0.3387758433818817, "learning_rate": 3.431502984866343e-05, "loss": 0.0122, "step": 62150 }, { "grad_norm": 0.32986438274383545, "learning_rate": 3.429933064119753e-05, "loss": 0.0259, "step": 62160 }, { "grad_norm": 0.24652957916259766, "learning_rate": 3.4283633150733255e-05, "loss": 0.0181, "step": 62170 }, { "grad_norm": 0.23489491641521454, "learning_rate": 3.4267937378987225e-05, "loss": 0.0225, "step": 62180 }, { "grad_norm": 0.18309877812862396, "learning_rate": 3.425224332767592e-05, "loss": 0.013, "step": 62190 }, { "grad_norm": 0.4778699576854706, "learning_rate": 3.4236550998515595e-05, "loss": 0.0194, "step": 62200 }, { "grad_norm": 0.25321805477142334, "learning_rate": 3.422086039322235e-05, "loss": 0.0136, "step": 62210 }, { "grad_norm": 0.35323458909988403, "learning_rate": 3.420517151351209e-05, "loss": 0.0193, "step": 62220 }, { "grad_norm": 0.11048249155282974, "learning_rate": 3.418948436110053e-05, "loss": 0.0139, "step": 62230 }, { "grad_norm": 0.25522318482398987, "learning_rate": 3.417379893770319e-05, "loss": 0.0271, "step": 62240 }, { "grad_norm": 0.21834483742713928, "learning_rate": 3.41581152450354e-05, "loss": 0.0121, "step": 62250 }, { "grad_norm": 0.1781749725341797, "learning_rate": 3.414243328481231e-05, "loss": 0.0096, "step": 62260 }, { "grad_norm": 0.2309529036283493, "learning_rate": 3.412675305874888e-05, "loss": 0.0261, "step": 62270 }, { "grad_norm": 0.1550447642803192, "learning_rate": 3.4111074568559876e-05, "loss": 0.0115, "step": 62280 }, { "grad_norm": 0.32711806893348694, "learning_rate": 3.4095397815959846e-05, "loss": 0.019, "step": 62290 }, { "grad_norm": 0.42057546973228455, "learning_rate": 3.407972280266322e-05, "loss": 0.0153, "step": 62300 }, { "grad_norm": 0.2528790533542633, "learning_rate": 3.4064049530384166e-05, "loss": 0.0159, "step": 62310 }, { "grad_norm": 0.1321418583393097, "learning_rate": 3.40483780008367e-05, "loss": 0.0135, "step": 62320 }, { "grad_norm": 0.119904063642025, "learning_rate": 3.4032708215734635e-05, "loss": 0.0208, "step": 62330 }, { "grad_norm": 0.25536224246025085, "learning_rate": 3.4017040176791584e-05, "loss": 0.0096, "step": 62340 }, { "grad_norm": 0.2339465618133545, "learning_rate": 3.4001373885720976e-05, "loss": 0.0241, "step": 62350 }, { "grad_norm": 0.3202953636646271, "learning_rate": 3.398570934423608e-05, "loss": 0.0185, "step": 62360 }, { "grad_norm": 0.10286567360162735, "learning_rate": 3.397004655404994e-05, "loss": 0.0168, "step": 62370 }, { "grad_norm": 0.3893669545650482, "learning_rate": 3.395438551687541e-05, "loss": 0.0195, "step": 62380 }, { "grad_norm": 0.09140836447477341, "learning_rate": 3.393872623442515e-05, "loss": 0.0165, "step": 62390 }, { "grad_norm": 0.15199065208435059, "learning_rate": 3.392306870841164e-05, "loss": 0.0087, "step": 62400 }, { "grad_norm": 0.28968289494514465, "learning_rate": 3.390741294054715e-05, "loss": 0.0148, "step": 62410 }, { "grad_norm": 0.10311630368232727, "learning_rate": 3.38917589325438e-05, "loss": 0.0141, "step": 62420 }, { "grad_norm": 0.2674315571784973, "learning_rate": 3.387610668611349e-05, "loss": 0.0126, "step": 62430 }, { "grad_norm": 0.35092824697494507, "learning_rate": 3.38604562029679e-05, "loss": 0.0128, "step": 62440 }, { "grad_norm": 0.19252097606658936, "learning_rate": 3.384480748481855e-05, "loss": 0.0155, "step": 62450 }, { "grad_norm": 0.2795545756816864, "learning_rate": 3.382916053337676e-05, "loss": 0.0166, "step": 62460 }, { "grad_norm": 0.37338775396347046, "learning_rate": 3.381351535035366e-05, "loss": 0.0177, "step": 62470 }, { "grad_norm": 0.1301800012588501, "learning_rate": 3.379787193746019e-05, "loss": 0.0153, "step": 62480 }, { "grad_norm": 0.06314299255609512, "learning_rate": 3.378223029640709e-05, "loss": 0.0119, "step": 62490 }, { "grad_norm": 0.1755225956439972, "learning_rate": 3.376659042890489e-05, "loss": 0.0101, "step": 62500 }, { "grad_norm": 0.11459009349346161, "learning_rate": 3.375095233666397e-05, "loss": 0.013, "step": 62510 }, { "grad_norm": 0.2196480631828308, "learning_rate": 3.373531602139447e-05, "loss": 0.0128, "step": 62520 }, { "grad_norm": 0.38179701566696167, "learning_rate": 3.371968148480635e-05, "loss": 0.0174, "step": 62530 }, { "grad_norm": 0.21835722029209137, "learning_rate": 3.370404872860937e-05, "loss": 0.0151, "step": 62540 }, { "grad_norm": 0.25926920771598816, "learning_rate": 3.3688417754513145e-05, "loss": 0.0164, "step": 62550 }, { "grad_norm": 0.1991327553987503, "learning_rate": 3.367278856422703e-05, "loss": 0.0183, "step": 62560 }, { "grad_norm": 0.1408180296421051, "learning_rate": 3.365716115946021e-05, "loss": 0.0137, "step": 62570 }, { "grad_norm": 0.1627274751663208, "learning_rate": 3.3641535541921666e-05, "loss": 0.0303, "step": 62580 }, { "grad_norm": 0.13688281178474426, "learning_rate": 3.36259117133202e-05, "loss": 0.0123, "step": 62590 }, { "grad_norm": 0.2859371304512024, "learning_rate": 3.36102896753644e-05, "loss": 0.0169, "step": 62600 }, { "grad_norm": 0.13714146614074707, "learning_rate": 3.35946694297627e-05, "loss": 0.0131, "step": 62610 }, { "grad_norm": 0.09406808763742447, "learning_rate": 3.357905097822328e-05, "loss": 0.0174, "step": 62620 }, { "grad_norm": 0.2914745509624481, "learning_rate": 3.356343432245416e-05, "loss": 0.0104, "step": 62630 }, { "grad_norm": 0.1882791370153427, "learning_rate": 3.354781946416315e-05, "loss": 0.028, "step": 62640 }, { "grad_norm": 0.17119194567203522, "learning_rate": 3.3532206405057864e-05, "loss": 0.0147, "step": 62650 }, { "grad_norm": 0.20472893118858337, "learning_rate": 3.351659514684573e-05, "loss": 0.0161, "step": 62660 }, { "grad_norm": 0.1661238819360733, "learning_rate": 3.350098569123395e-05, "loss": 0.0129, "step": 62670 }, { "grad_norm": 0.1962934285402298, "learning_rate": 3.348537803992958e-05, "loss": 0.0113, "step": 62680 }, { "grad_norm": 0.4600204825401306, "learning_rate": 3.346977219463945e-05, "loss": 0.015, "step": 62690 }, { "grad_norm": 0.13227617740631104, "learning_rate": 3.345416815707017e-05, "loss": 0.0153, "step": 62700 }, { "grad_norm": 0.14419257640838623, "learning_rate": 3.3438565928928185e-05, "loss": 0.0121, "step": 62710 }, { "grad_norm": 0.2534578740596771, "learning_rate": 3.3422965511919736e-05, "loss": 0.0148, "step": 62720 }, { "grad_norm": 0.11688139289617538, "learning_rate": 3.3407366907750846e-05, "loss": 0.0129, "step": 62730 }, { "grad_norm": 0.12355183064937592, "learning_rate": 3.339177011812737e-05, "loss": 0.0188, "step": 62740 }, { "grad_norm": 0.18262754380702972, "learning_rate": 3.3376175144754935e-05, "loss": 0.0207, "step": 62750 }, { "grad_norm": 0.21323515474796295, "learning_rate": 3.3360581989339004e-05, "loss": 0.0219, "step": 62760 }, { "grad_norm": 0.07979585230350494, "learning_rate": 3.3344990653584804e-05, "loss": 0.0157, "step": 62770 }, { "grad_norm": 0.15012934803962708, "learning_rate": 3.332940113919739e-05, "loss": 0.017, "step": 62780 }, { "grad_norm": 0.11712543666362762, "learning_rate": 3.331381344788161e-05, "loss": 0.0121, "step": 62790 }, { "grad_norm": 0.1375395506620407, "learning_rate": 3.3298227581342076e-05, "loss": 0.0154, "step": 62800 }, { "grad_norm": 0.3173829913139343, "learning_rate": 3.328264354128329e-05, "loss": 0.0231, "step": 62810 }, { "grad_norm": 0.09356889873743057, "learning_rate": 3.3267061329409465e-05, "loss": 0.024, "step": 62820 }, { "grad_norm": 0.1273597925901413, "learning_rate": 3.325148094742466e-05, "loss": 0.0114, "step": 62830 }, { "grad_norm": 0.2710385322570801, "learning_rate": 3.323590239703271e-05, "loss": 0.0171, "step": 62840 }, { "grad_norm": 0.10142796486616135, "learning_rate": 3.322032567993728e-05, "loss": 0.021, "step": 62850 }, { "grad_norm": 0.2834477722644806, "learning_rate": 3.3204750797841774e-05, "loss": 0.0193, "step": 62860 }, { "grad_norm": 0.19766372442245483, "learning_rate": 3.31891777524495e-05, "loss": 0.0159, "step": 62870 }, { "grad_norm": 0.1000954881310463, "learning_rate": 3.317360654546347e-05, "loss": 0.0156, "step": 62880 }, { "grad_norm": 0.07715237140655518, "learning_rate": 3.3158037178586534e-05, "loss": 0.0125, "step": 62890 }, { "grad_norm": 0.2085195630788803, "learning_rate": 3.3142469653521316e-05, "loss": 0.0175, "step": 62900 }, { "grad_norm": 0.1303582787513733, "learning_rate": 3.3126903971970284e-05, "loss": 0.0112, "step": 62910 }, { "grad_norm": 0.5693607926368713, "learning_rate": 3.311134013563564e-05, "loss": 0.0181, "step": 62920 }, { "grad_norm": 0.2664780020713806, "learning_rate": 3.309577814621947e-05, "loss": 0.0182, "step": 62930 }, { "grad_norm": 0.19844192266464233, "learning_rate": 3.308021800542359e-05, "loss": 0.0129, "step": 62940 }, { "grad_norm": 0.4490959346294403, "learning_rate": 3.306465971494963e-05, "loss": 0.0183, "step": 62950 }, { "grad_norm": 0.275546669960022, "learning_rate": 3.304910327649901e-05, "loss": 0.015, "step": 62960 }, { "grad_norm": 0.21222157776355743, "learning_rate": 3.303354869177297e-05, "loss": 0.0134, "step": 62970 }, { "grad_norm": 0.1053369864821434, "learning_rate": 3.301799596247255e-05, "loss": 0.0138, "step": 62980 }, { "grad_norm": 0.08664596825838089, "learning_rate": 3.300244509029854e-05, "loss": 0.0105, "step": 62990 }, { "grad_norm": 0.2015983909368515, "learning_rate": 3.2986896076951585e-05, "loss": 0.0133, "step": 63000 }, { "grad_norm": 0.23271849751472473, "learning_rate": 3.297134892413211e-05, "loss": 0.022, "step": 63010 }, { "grad_norm": 0.0764991044998169, "learning_rate": 3.295580363354031e-05, "loss": 0.0177, "step": 63020 }, { "grad_norm": 0.40444526076316833, "learning_rate": 3.2940260206876194e-05, "loss": 0.0164, "step": 63030 }, { "grad_norm": 0.20827654004096985, "learning_rate": 3.292471864583958e-05, "loss": 0.0131, "step": 63040 }, { "grad_norm": 0.1299179196357727, "learning_rate": 3.290917895213004e-05, "loss": 0.0146, "step": 63050 }, { "grad_norm": 0.24015122652053833, "learning_rate": 3.289364112744701e-05, "loss": 0.0184, "step": 63060 }, { "grad_norm": 0.4038735330104828, "learning_rate": 3.287810517348967e-05, "loss": 0.0262, "step": 63070 }, { "grad_norm": 0.14628010988235474, "learning_rate": 3.2862571091957014e-05, "loss": 0.0213, "step": 63080 }, { "grad_norm": 0.5295348167419434, "learning_rate": 3.284703888454781e-05, "loss": 0.0241, "step": 63090 }, { "grad_norm": 0.41877657175064087, "learning_rate": 3.2831508552960646e-05, "loss": 0.0201, "step": 63100 }, { "grad_norm": 0.22110480070114136, "learning_rate": 3.281598009889387e-05, "loss": 0.0218, "step": 63110 }, { "grad_norm": 0.09948108345270157, "learning_rate": 3.28004535240457e-05, "loss": 0.016, "step": 63120 }, { "grad_norm": 0.18462662398815155, "learning_rate": 3.278492883011407e-05, "loss": 0.0112, "step": 63130 }, { "grad_norm": 0.24323581159114838, "learning_rate": 3.276940601879675e-05, "loss": 0.0168, "step": 63140 }, { "grad_norm": 0.14607255160808563, "learning_rate": 3.275388509179127e-05, "loss": 0.0152, "step": 63150 }, { "grad_norm": 0.143472358584404, "learning_rate": 3.2738366050794985e-05, "loss": 0.0159, "step": 63160 }, { "grad_norm": 0.18954186141490936, "learning_rate": 3.2722848897505034e-05, "loss": 0.0134, "step": 63170 }, { "grad_norm": 0.40211760997772217, "learning_rate": 3.2707333633618356e-05, "loss": 0.011, "step": 63180 }, { "grad_norm": 0.14867550134658813, "learning_rate": 3.269182026083167e-05, "loss": 0.0169, "step": 63190 }, { "grad_norm": 0.09967567026615143, "learning_rate": 3.2676308780841516e-05, "loss": 0.0265, "step": 63200 }, { "grad_norm": 0.18624503910541534, "learning_rate": 3.266079919534416e-05, "loss": 0.0129, "step": 63210 }, { "grad_norm": 0.0844087079167366, "learning_rate": 3.264529150603575e-05, "loss": 0.0188, "step": 63220 }, { "grad_norm": 0.4030616581439972, "learning_rate": 3.262978571461217e-05, "loss": 0.0154, "step": 63230 }, { "grad_norm": 0.14276453852653503, "learning_rate": 3.26142818227691e-05, "loss": 0.0107, "step": 63240 }, { "grad_norm": 0.1873786747455597, "learning_rate": 3.259877983220203e-05, "loss": 0.0182, "step": 63250 }, { "grad_norm": 0.0818106159567833, "learning_rate": 3.2583279744606246e-05, "loss": 0.0202, "step": 63260 }, { "grad_norm": 0.10644800215959549, "learning_rate": 3.25677815616768e-05, "loss": 0.0142, "step": 63270 }, { "grad_norm": 0.12989619374275208, "learning_rate": 3.255228528510856e-05, "loss": 0.0169, "step": 63280 }, { "grad_norm": 0.22945575416088104, "learning_rate": 3.253679091659617e-05, "loss": 0.018, "step": 63290 }, { "grad_norm": 0.2703998386859894, "learning_rate": 3.252129845783404e-05, "loss": 0.0188, "step": 63300 }, { "grad_norm": 0.2972089946269989, "learning_rate": 3.2505807910516456e-05, "loss": 0.0102, "step": 63310 }, { "grad_norm": 0.20613116025924683, "learning_rate": 3.249031927633742e-05, "loss": 0.0172, "step": 63320 }, { "grad_norm": 0.27101171016693115, "learning_rate": 3.2474832556990737e-05, "loss": 0.0135, "step": 63330 }, { "grad_norm": 0.18058672547340393, "learning_rate": 3.245934775417002e-05, "loss": 0.0221, "step": 63340 }, { "grad_norm": 0.28800755739212036, "learning_rate": 3.244386486956866e-05, "loss": 0.0174, "step": 63350 }, { "grad_norm": 0.09764939546585083, "learning_rate": 3.2428383904879846e-05, "loss": 0.012, "step": 63360 }, { "grad_norm": 0.20847441256046295, "learning_rate": 3.241290486179652e-05, "loss": 0.0326, "step": 63370 }, { "grad_norm": 0.22150462865829468, "learning_rate": 3.239742774201151e-05, "loss": 0.0165, "step": 63380 }, { "grad_norm": 0.06901945918798447, "learning_rate": 3.238195254721733e-05, "loss": 0.012, "step": 63390 }, { "grad_norm": 0.10035935789346695, "learning_rate": 3.236647927910633e-05, "loss": 0.0108, "step": 63400 }, { "grad_norm": 0.055303238332271576, "learning_rate": 3.2351007939370646e-05, "loss": 0.0117, "step": 63410 }, { "grad_norm": 0.12049134075641632, "learning_rate": 3.2335538529702184e-05, "loss": 0.0101, "step": 63420 }, { "grad_norm": 0.14218541979789734, "learning_rate": 3.232007105179268e-05, "loss": 0.0166, "step": 63430 }, { "grad_norm": 0.36774399876594543, "learning_rate": 3.230460550733363e-05, "loss": 0.0139, "step": 63440 }, { "grad_norm": 0.2144308239221573, "learning_rate": 3.2289141898016315e-05, "loss": 0.0095, "step": 63450 }, { "grad_norm": 0.206930473446846, "learning_rate": 3.227368022553181e-05, "loss": 0.0175, "step": 63460 }, { "grad_norm": 0.15918804705142975, "learning_rate": 3.225822049157099e-05, "loss": 0.0138, "step": 63470 }, { "grad_norm": 0.38272324204444885, "learning_rate": 3.2242762697824494e-05, "loss": 0.0206, "step": 63480 }, { "grad_norm": 0.2624274790287018, "learning_rate": 3.222730684598279e-05, "loss": 0.0197, "step": 63490 }, { "grad_norm": 0.246846005320549, "learning_rate": 3.221185293773605e-05, "loss": 0.0244, "step": 63500 }, { "grad_norm": 0.18232345581054688, "learning_rate": 3.2196400974774354e-05, "loss": 0.0127, "step": 63510 }, { "grad_norm": 0.12711137533187866, "learning_rate": 3.2180950958787485e-05, "loss": 0.0218, "step": 63520 }, { "grad_norm": 0.18563911318778992, "learning_rate": 3.2165502891465024e-05, "loss": 0.0215, "step": 63530 }, { "grad_norm": 0.17332464456558228, "learning_rate": 3.215005677449636e-05, "loss": 0.0118, "step": 63540 }, { "grad_norm": 0.1639091670513153, "learning_rate": 3.2134612609570644e-05, "loss": 0.0111, "step": 63550 }, { "grad_norm": 0.17641444504261017, "learning_rate": 3.211917039837682e-05, "loss": 0.0128, "step": 63560 }, { "grad_norm": 0.25621190667152405, "learning_rate": 3.210373014260365e-05, "loss": 0.0332, "step": 63570 }, { "grad_norm": 0.1091289147734642, "learning_rate": 3.208829184393965e-05, "loss": 0.0117, "step": 63580 }, { "grad_norm": 0.1416129171848297, "learning_rate": 3.207285550407312e-05, "loss": 0.0168, "step": 63590 }, { "grad_norm": 0.25127971172332764, "learning_rate": 3.205742112469217e-05, "loss": 0.0087, "step": 63600 }, { "grad_norm": 0.25364455580711365, "learning_rate": 3.204198870748465e-05, "loss": 0.0183, "step": 63610 }, { "grad_norm": 0.20307382941246033, "learning_rate": 3.202655825413824e-05, "loss": 0.0173, "step": 63620 }, { "grad_norm": 0.25414901971817017, "learning_rate": 3.2011129766340416e-05, "loss": 0.0244, "step": 63630 }, { "grad_norm": 0.3793807923793793, "learning_rate": 3.1995703245778384e-05, "loss": 0.0149, "step": 63640 }, { "grad_norm": 0.07739521563053131, "learning_rate": 3.1980278694139186e-05, "loss": 0.0141, "step": 63650 }, { "grad_norm": 0.1270720660686493, "learning_rate": 3.196485611310962e-05, "loss": 0.0129, "step": 63660 }, { "grad_norm": 0.22798988223075867, "learning_rate": 3.1949435504376257e-05, "loss": 0.0183, "step": 63670 }, { "grad_norm": 0.09318063408136368, "learning_rate": 3.1934016869625505e-05, "loss": 0.0097, "step": 63680 }, { "grad_norm": 0.15427471697330475, "learning_rate": 3.191860021054349e-05, "loss": 0.0232, "step": 63690 }, { "grad_norm": 0.15529948472976685, "learning_rate": 3.190318552881618e-05, "loss": 0.0215, "step": 63700 }, { "grad_norm": 0.220398411154747, "learning_rate": 3.1887772826129285e-05, "loss": 0.0179, "step": 63710 }, { "grad_norm": 0.2600870728492737, "learning_rate": 3.1872362104168335e-05, "loss": 0.0187, "step": 63720 }, { "grad_norm": 0.12154809385538101, "learning_rate": 3.185695336461861e-05, "loss": 0.0125, "step": 63730 }, { "grad_norm": 0.18028020858764648, "learning_rate": 3.184154660916519e-05, "loss": 0.0186, "step": 63740 }, { "grad_norm": 0.20365017652511597, "learning_rate": 3.1826141839492906e-05, "loss": 0.02, "step": 63750 }, { "grad_norm": 0.2866373360157013, "learning_rate": 3.181073905728644e-05, "loss": 0.0161, "step": 63760 }, { "grad_norm": 0.06983768939971924, "learning_rate": 3.1795338264230214e-05, "loss": 0.0118, "step": 63770 }, { "grad_norm": 0.20694737136363983, "learning_rate": 3.1779939462008426e-05, "loss": 0.0263, "step": 63780 }, { "grad_norm": 0.16734887659549713, "learning_rate": 3.1764542652305066e-05, "loss": 0.0163, "step": 63790 }, { "grad_norm": 0.07920293509960175, "learning_rate": 3.17491478368039e-05, "loss": 0.0147, "step": 63800 }, { "grad_norm": 0.21606189012527466, "learning_rate": 3.173375501718847e-05, "loss": 0.0224, "step": 63810 }, { "grad_norm": 0.14693622291088104, "learning_rate": 3.171836419514215e-05, "loss": 0.0151, "step": 63820 }, { "grad_norm": 0.2263779193162918, "learning_rate": 3.170297537234803e-05, "loss": 0.0192, "step": 63830 }, { "grad_norm": 0.17002542316913605, "learning_rate": 3.168758855048901e-05, "loss": 0.0133, "step": 63840 }, { "grad_norm": 0.2455761581659317, "learning_rate": 3.1672203731247785e-05, "loss": 0.0152, "step": 63850 }, { "grad_norm": 0.17942173779010773, "learning_rate": 3.16568209163068e-05, "loss": 0.0165, "step": 63860 }, { "grad_norm": 0.1411939263343811, "learning_rate": 3.1641440107348295e-05, "loss": 0.0145, "step": 63870 }, { "grad_norm": 0.3893905282020569, "learning_rate": 3.162606130605429e-05, "loss": 0.0154, "step": 63880 }, { "grad_norm": 0.07469858229160309, "learning_rate": 3.16106845141066e-05, "loss": 0.0118, "step": 63890 }, { "grad_norm": 0.2420661449432373, "learning_rate": 3.159530973318682e-05, "loss": 0.0124, "step": 63900 }, { "grad_norm": 0.16150076687335968, "learning_rate": 3.1579936964976286e-05, "loss": 0.0102, "step": 63910 }, { "grad_norm": 0.18772689998149872, "learning_rate": 3.156456621115615e-05, "loss": 0.0228, "step": 63920 }, { "grad_norm": 0.16818083822727203, "learning_rate": 3.154919747340734e-05, "loss": 0.0123, "step": 63930 }, { "grad_norm": 0.27013668417930603, "learning_rate": 3.153383075341054e-05, "loss": 0.0149, "step": 63940 }, { "grad_norm": 0.1445448100566864, "learning_rate": 3.1518466052846264e-05, "loss": 0.0195, "step": 63950 }, { "grad_norm": 0.1988297700881958, "learning_rate": 3.150310337339474e-05, "loss": 0.014, "step": 63960 }, { "grad_norm": 0.11662528663873672, "learning_rate": 3.148774271673603e-05, "loss": 0.0107, "step": 63970 }, { "grad_norm": 0.1650339961051941, "learning_rate": 3.147238408454995e-05, "loss": 0.0159, "step": 63980 }, { "grad_norm": 0.24127377569675446, "learning_rate": 3.1457027478516086e-05, "loss": 0.0151, "step": 63990 }, { "grad_norm": 0.21987944841384888, "learning_rate": 3.144167290031382e-05, "loss": 0.0127, "step": 64000 }, { "grad_norm": 0.3745191693305969, "learning_rate": 3.142632035162227e-05, "loss": 0.0162, "step": 64010 }, { "grad_norm": 0.2119891494512558, "learning_rate": 3.141096983412044e-05, "loss": 0.0116, "step": 64020 }, { "grad_norm": 0.16304922103881836, "learning_rate": 3.139562134948699e-05, "loss": 0.012, "step": 64030 }, { "grad_norm": 0.13249057531356812, "learning_rate": 3.1380274899400424e-05, "loss": 0.0125, "step": 64040 }, { "grad_norm": 0.13170704245567322, "learning_rate": 3.1364930485539e-05, "loss": 0.0132, "step": 64050 }, { "grad_norm": 0.19366730749607086, "learning_rate": 3.134958810958076e-05, "loss": 0.0113, "step": 64060 }, { "grad_norm": 0.19947758316993713, "learning_rate": 3.13342477732035e-05, "loss": 0.0123, "step": 64070 }, { "grad_norm": 0.25567537546157837, "learning_rate": 3.131890947808487e-05, "loss": 0.0126, "step": 64080 }, { "grad_norm": 0.3435715436935425, "learning_rate": 3.13035732259022e-05, "loss": 0.0129, "step": 64090 }, { "grad_norm": 0.09434811770915985, "learning_rate": 3.1288239018332656e-05, "loss": 0.0127, "step": 64100 }, { "grad_norm": 0.18368782103061676, "learning_rate": 3.1272906857053164e-05, "loss": 0.015, "step": 64110 }, { "grad_norm": 0.3433675467967987, "learning_rate": 3.125757674374042e-05, "loss": 0.02, "step": 64120 }, { "grad_norm": 0.30965206027030945, "learning_rate": 3.124224868007088e-05, "loss": 0.019, "step": 64130 }, { "grad_norm": 0.10674426704645157, "learning_rate": 3.1226922667720854e-05, "loss": 0.0107, "step": 64140 }, { "grad_norm": 0.2291603982448578, "learning_rate": 3.121159870836633e-05, "loss": 0.021, "step": 64150 }, { "grad_norm": 0.1077752336859703, "learning_rate": 3.119627680368312e-05, "loss": 0.009, "step": 64160 }, { "grad_norm": 0.6350327134132385, "learning_rate": 3.1180956955346816e-05, "loss": 0.023, "step": 64170 }, { "grad_norm": 0.1799759566783905, "learning_rate": 3.116563916503276e-05, "loss": 0.0176, "step": 64180 }, { "grad_norm": 0.15646415948867798, "learning_rate": 3.1150323434416095e-05, "loss": 0.0198, "step": 64190 }, { "grad_norm": 0.21004128456115723, "learning_rate": 3.1135009765171696e-05, "loss": 0.008, "step": 64200 }, { "grad_norm": 0.14928479492664337, "learning_rate": 3.111969815897429e-05, "loss": 0.0161, "step": 64210 }, { "grad_norm": 0.14483770728111267, "learning_rate": 3.1104388617498295e-05, "loss": 0.0105, "step": 64220 }, { "grad_norm": 0.20743530988693237, "learning_rate": 3.1089081142417966e-05, "loss": 0.0161, "step": 64230 }, { "grad_norm": 0.06718282401561737, "learning_rate": 3.1073775735407285e-05, "loss": 0.0123, "step": 64240 }, { "grad_norm": 0.28021642565727234, "learning_rate": 3.1058472398140035e-05, "loss": 0.0097, "step": 64250 }, { "grad_norm": 0.08055242896080017, "learning_rate": 3.1043171132289736e-05, "loss": 0.0124, "step": 64260 }, { "grad_norm": 0.11340873688459396, "learning_rate": 3.1027871939529774e-05, "loss": 0.0143, "step": 64270 }, { "grad_norm": 0.26418593525886536, "learning_rate": 3.1012574821533214e-05, "loss": 0.0223, "step": 64280 }, { "grad_norm": 0.31197142601013184, "learning_rate": 3.099727977997291e-05, "loss": 0.0163, "step": 64290 }, { "grad_norm": 0.10572279244661331, "learning_rate": 3.098198681652154e-05, "loss": 0.0165, "step": 64300 }, { "grad_norm": 0.1269795000553131, "learning_rate": 3.0966695932851475e-05, "loss": 0.0131, "step": 64310 }, { "grad_norm": 0.1409471184015274, "learning_rate": 3.0951407130634926e-05, "loss": 0.0149, "step": 64320 }, { "grad_norm": 0.18529506027698517, "learning_rate": 3.0936120411543864e-05, "loss": 0.0201, "step": 64330 }, { "grad_norm": 0.18417982757091522, "learning_rate": 3.0920835777250015e-05, "loss": 0.0203, "step": 64340 }, { "grad_norm": 0.10867980867624283, "learning_rate": 3.090555322942488e-05, "loss": 0.014, "step": 64350 }, { "grad_norm": 0.08653608709573746, "learning_rate": 3.089027276973973e-05, "loss": 0.0208, "step": 64360 }, { "grad_norm": 0.11650586128234863, "learning_rate": 3.0874994399865615e-05, "loss": 0.0137, "step": 64370 }, { "grad_norm": 0.15132442116737366, "learning_rate": 3.085971812147337e-05, "loss": 0.0136, "step": 64380 }, { "grad_norm": 0.1414966583251953, "learning_rate": 3.0844443936233557e-05, "loss": 0.0134, "step": 64390 }, { "grad_norm": 0.15295635163784027, "learning_rate": 3.082917184581657e-05, "loss": 0.0187, "step": 64400 }, { "grad_norm": 0.1909061074256897, "learning_rate": 3.0813901851892513e-05, "loss": 0.0088, "step": 64410 }, { "grad_norm": 0.2285868525505066, "learning_rate": 3.0798633956131316e-05, "loss": 0.0242, "step": 64420 }, { "grad_norm": 0.1823442280292511, "learning_rate": 3.078336816020264e-05, "loss": 0.0155, "step": 64430 }, { "grad_norm": 0.15959686040878296, "learning_rate": 3.0768104465775935e-05, "loss": 0.0142, "step": 64440 }, { "grad_norm": 0.2674638330936432, "learning_rate": 3.0752842874520396e-05, "loss": 0.0156, "step": 64450 }, { "grad_norm": 0.17595134675502777, "learning_rate": 3.073758338810503e-05, "loss": 0.015, "step": 64460 }, { "grad_norm": 0.18588823080062866, "learning_rate": 3.072232600819859e-05, "loss": 0.0135, "step": 64470 }, { "grad_norm": 0.2873377203941345, "learning_rate": 3.070707073646959e-05, "loss": 0.0214, "step": 64480 }, { "grad_norm": 0.13679565489292145, "learning_rate": 3.069181757458633e-05, "loss": 0.0112, "step": 64490 }, { "grad_norm": 0.16083090007305145, "learning_rate": 3.067656652421687e-05, "loss": 0.0118, "step": 64500 }, { "grad_norm": 0.32113566994667053, "learning_rate": 3.0661317587029025e-05, "loss": 0.0234, "step": 64510 }, { "grad_norm": 0.11411619186401367, "learning_rate": 3.064607076469043e-05, "loss": 0.0164, "step": 64520 }, { "grad_norm": 0.1112305074930191, "learning_rate": 3.063082605886845e-05, "loss": 0.017, "step": 64530 }, { "grad_norm": 0.10926289111375809, "learning_rate": 3.061558347123019e-05, "loss": 0.0162, "step": 64540 }, { "grad_norm": 0.2795121669769287, "learning_rate": 3.060034300344259e-05, "loss": 0.0137, "step": 64550 }, { "grad_norm": 0.17761053144931793, "learning_rate": 3.058510465717231e-05, "loss": 0.0137, "step": 64560 }, { "grad_norm": 0.17080926895141602, "learning_rate": 3.0569868434085804e-05, "loss": 0.0172, "step": 64570 }, { "grad_norm": 0.05483529344201088, "learning_rate": 3.055463433584924e-05, "loss": 0.0115, "step": 64580 }, { "grad_norm": 0.2459132820367813, "learning_rate": 3.0539402364128656e-05, "loss": 0.0149, "step": 64590 }, { "grad_norm": 0.18855050206184387, "learning_rate": 3.052417252058977e-05, "loss": 0.0117, "step": 64600 }, { "grad_norm": 0.16624298691749573, "learning_rate": 3.0508944806898092e-05, "loss": 0.0232, "step": 64610 }, { "grad_norm": 0.08363623172044754, "learning_rate": 3.04937192247189e-05, "loss": 0.0105, "step": 64620 }, { "grad_norm": 0.07992592453956604, "learning_rate": 3.047849577571725e-05, "loss": 0.0148, "step": 64630 }, { "grad_norm": 0.3751591444015503, "learning_rate": 3.046327446155795e-05, "loss": 0.0176, "step": 64640 }, { "grad_norm": 0.23321713507175446, "learning_rate": 3.0448055283905587e-05, "loss": 0.0084, "step": 64650 }, { "grad_norm": 0.2229202687740326, "learning_rate": 3.0432838244424494e-05, "loss": 0.0135, "step": 64660 }, { "grad_norm": 0.12750430405139923, "learning_rate": 3.04176233447788e-05, "loss": 0.0113, "step": 64670 }, { "grad_norm": 0.16809768974781036, "learning_rate": 3.040241058663238e-05, "loss": 0.0151, "step": 64680 }, { "grad_norm": 0.2978714108467102, "learning_rate": 3.038719997164887e-05, "loss": 0.0258, "step": 64690 }, { "grad_norm": 0.22777129709720612, "learning_rate": 3.0371991501491683e-05, "loss": 0.0183, "step": 64700 }, { "grad_norm": 0.18897411227226257, "learning_rate": 3.0356785177823975e-05, "loss": 0.012, "step": 64710 }, { "grad_norm": 0.32030045986175537, "learning_rate": 3.0341581002308726e-05, "loss": 0.0168, "step": 64720 }, { "grad_norm": 0.2896764576435089, "learning_rate": 3.0326378976608627e-05, "loss": 0.0168, "step": 64730 }, { "grad_norm": 0.06881643831729889, "learning_rate": 3.0311179102386146e-05, "loss": 0.0184, "step": 64740 }, { "grad_norm": 0.14038752019405365, "learning_rate": 3.0295981381303513e-05, "loss": 0.0109, "step": 64750 }, { "grad_norm": 0.4707188904285431, "learning_rate": 3.0280785815022727e-05, "loss": 0.0135, "step": 64760 }, { "grad_norm": 0.25252577662467957, "learning_rate": 3.0265592405205544e-05, "loss": 0.0116, "step": 64770 }, { "grad_norm": 0.2104351967573166, "learning_rate": 3.0250401153513517e-05, "loss": 0.0157, "step": 64780 }, { "grad_norm": 0.15098239481449127, "learning_rate": 3.0235212061607936e-05, "loss": 0.0226, "step": 64790 }, { "grad_norm": 0.23271119594573975, "learning_rate": 3.0220025131149842e-05, "loss": 0.0129, "step": 64800 }, { "grad_norm": 0.23846852779388428, "learning_rate": 3.020484036380007e-05, "loss": 0.013, "step": 64810 }, { "grad_norm": 0.2580239772796631, "learning_rate": 3.0189657761219188e-05, "loss": 0.0164, "step": 64820 }, { "grad_norm": 0.35461297631263733, "learning_rate": 3.017447732506753e-05, "loss": 0.0173, "step": 64830 }, { "grad_norm": 0.15654747188091278, "learning_rate": 3.015929905700524e-05, "loss": 0.0124, "step": 64840 }, { "grad_norm": 0.20344574749469757, "learning_rate": 3.014412295869218e-05, "loss": 0.0097, "step": 64850 }, { "grad_norm": 0.5304582118988037, "learning_rate": 3.0128949031787977e-05, "loss": 0.0178, "step": 64860 }, { "grad_norm": 0.1028549075126648, "learning_rate": 3.0113777277952022e-05, "loss": 0.016, "step": 64870 }, { "grad_norm": 0.4043086767196655, "learning_rate": 3.0098607698843496e-05, "loss": 0.0147, "step": 64880 }, { "grad_norm": 0.26726916432380676, "learning_rate": 3.0083440296121312e-05, "loss": 0.0168, "step": 64890 }, { "grad_norm": 0.34734907746315, "learning_rate": 3.0068275071444142e-05, "loss": 0.014, "step": 64900 }, { "grad_norm": 0.29778942465782166, "learning_rate": 3.0053112026470442e-05, "loss": 0.0244, "step": 64910 }, { "grad_norm": 0.3359310030937195, "learning_rate": 3.0037951162858434e-05, "loss": 0.012, "step": 64920 }, { "grad_norm": 0.13163550198078156, "learning_rate": 3.002279248226607e-05, "loss": 0.0184, "step": 64930 }, { "grad_norm": 0.17041775584220886, "learning_rate": 3.000763598635109e-05, "loss": 0.0097, "step": 64940 }, { "grad_norm": 0.4735313951969147, "learning_rate": 2.999248167677098e-05, "loss": 0.0175, "step": 64950 }, { "grad_norm": 0.25516197085380554, "learning_rate": 2.9977329555182976e-05, "loss": 0.0181, "step": 64960 }, { "grad_norm": 0.1565190553665161, "learning_rate": 2.9962179623244124e-05, "loss": 0.0157, "step": 64970 }, { "grad_norm": 0.16621974110603333, "learning_rate": 2.9947031882611175e-05, "loss": 0.0173, "step": 64980 }, { "grad_norm": 0.12311800569295883, "learning_rate": 2.9931886334940685e-05, "loss": 0.0121, "step": 64990 }, { "grad_norm": 0.28834906220436096, "learning_rate": 2.9916742981888923e-05, "loss": 0.0193, "step": 65000 }, { "grad_norm": 0.09086743742227554, "learning_rate": 2.990160182511196e-05, "loss": 0.0182, "step": 65010 }, { "grad_norm": 0.2516980767250061, "learning_rate": 2.9886462866265574e-05, "loss": 0.015, "step": 65020 }, { "grad_norm": 0.19526846706867218, "learning_rate": 2.9871326107005392e-05, "loss": 0.016, "step": 65030 }, { "grad_norm": 0.2872597277164459, "learning_rate": 2.9856191548986735e-05, "loss": 0.0164, "step": 65040 }, { "grad_norm": 0.13377602398395538, "learning_rate": 2.9841059193864672e-05, "loss": 0.0151, "step": 65050 }, { "grad_norm": 0.1303919404745102, "learning_rate": 2.982592904329407e-05, "loss": 0.0185, "step": 65060 }, { "grad_norm": 0.25216811895370483, "learning_rate": 2.981080109892954e-05, "loss": 0.0144, "step": 65070 }, { "grad_norm": 0.10197847336530685, "learning_rate": 2.9795675362425434e-05, "loss": 0.0101, "step": 65080 }, { "grad_norm": 0.2568778395652771, "learning_rate": 2.97805518354359e-05, "loss": 0.0161, "step": 65090 }, { "grad_norm": 0.14698342978954315, "learning_rate": 2.9765430519614825e-05, "loss": 0.0094, "step": 65100 }, { "grad_norm": 0.26242518424987793, "learning_rate": 2.975031141661584e-05, "loss": 0.0212, "step": 65110 }, { "grad_norm": 0.32894137501716614, "learning_rate": 2.9735194528092347e-05, "loss": 0.0191, "step": 65120 }, { "grad_norm": 0.18557515740394592, "learning_rate": 2.9720079855697525e-05, "loss": 0.0147, "step": 65130 }, { "grad_norm": 0.08418715745210648, "learning_rate": 2.970496740108427e-05, "loss": 0.0183, "step": 65140 }, { "grad_norm": 0.21115244925022125, "learning_rate": 2.968985716590527e-05, "loss": 0.0122, "step": 65150 }, { "grad_norm": 0.273857980966568, "learning_rate": 2.9674749151812953e-05, "loss": 0.0132, "step": 65160 }, { "grad_norm": 0.15079733729362488, "learning_rate": 2.965964336045951e-05, "loss": 0.0115, "step": 65170 }, { "grad_norm": 0.19785401225090027, "learning_rate": 2.9644539793496894e-05, "loss": 0.0133, "step": 65180 }, { "grad_norm": 0.2512151300907135, "learning_rate": 2.9629438452576807e-05, "loss": 0.0141, "step": 65190 }, { "grad_norm": 0.10070251673460007, "learning_rate": 2.9614339339350698e-05, "loss": 0.0107, "step": 65200 }, { "grad_norm": 0.11404259502887726, "learning_rate": 2.959924245546979e-05, "loss": 0.0158, "step": 65210 }, { "grad_norm": 0.18322592973709106, "learning_rate": 2.9584147802585038e-05, "loss": 0.0153, "step": 65220 }, { "grad_norm": 0.34077516198158264, "learning_rate": 2.956905538234721e-05, "loss": 0.0122, "step": 65230 }, { "grad_norm": 0.28400999307632446, "learning_rate": 2.9553965196406764e-05, "loss": 0.0106, "step": 65240 }, { "grad_norm": 0.14511267840862274, "learning_rate": 2.9538877246413943e-05, "loss": 0.0156, "step": 65250 }, { "grad_norm": 0.3351336717605591, "learning_rate": 2.952379153401874e-05, "loss": 0.0197, "step": 65260 }, { "grad_norm": 0.12432761490345001, "learning_rate": 2.9508708060870914e-05, "loss": 0.0111, "step": 65270 }, { "grad_norm": 0.08452028036117554, "learning_rate": 2.9493626828619948e-05, "loss": 0.0075, "step": 65280 }, { "grad_norm": 0.18074427545070648, "learning_rate": 2.9478547838915137e-05, "loss": 0.0136, "step": 65290 }, { "grad_norm": 0.2293042540550232, "learning_rate": 2.946347109340548e-05, "loss": 0.0156, "step": 65300 }, { "grad_norm": 0.36309579014778137, "learning_rate": 2.944839659373976e-05, "loss": 0.0204, "step": 65310 }, { "grad_norm": 0.4138367772102356, "learning_rate": 2.943332434156647e-05, "loss": 0.022, "step": 65320 }, { "grad_norm": 0.11717607080936432, "learning_rate": 2.9418254338533912e-05, "loss": 0.0168, "step": 65330 }, { "grad_norm": 0.1644350290298462, "learning_rate": 2.940318658629011e-05, "loss": 0.0189, "step": 65340 }, { "grad_norm": 0.25495317578315735, "learning_rate": 2.938812108648287e-05, "loss": 0.0179, "step": 65350 }, { "grad_norm": 0.16572506725788116, "learning_rate": 2.937305784075972e-05, "loss": 0.0133, "step": 65360 }, { "grad_norm": 0.13504479825496674, "learning_rate": 2.9357996850767937e-05, "loss": 0.0143, "step": 65370 }, { "grad_norm": 0.2542774975299835, "learning_rate": 2.9342938118154595e-05, "loss": 0.0126, "step": 65380 }, { "grad_norm": 0.12687639892101288, "learning_rate": 2.9327881644566485e-05, "loss": 0.0123, "step": 65390 }, { "grad_norm": 0.117435522377491, "learning_rate": 2.9312827431650157e-05, "loss": 0.0147, "step": 65400 }, { "grad_norm": 0.8244063258171082, "learning_rate": 2.92977754810519e-05, "loss": 0.016, "step": 65410 }, { "grad_norm": 0.39907506108283997, "learning_rate": 2.928272579441781e-05, "loss": 0.0178, "step": 65420 }, { "grad_norm": 0.09439399093389511, "learning_rate": 2.9267678373393676e-05, "loss": 0.0184, "step": 65430 }, { "grad_norm": 0.1940382868051529, "learning_rate": 2.925263321962507e-05, "loss": 0.0172, "step": 65440 }, { "grad_norm": 0.1525871604681015, "learning_rate": 2.9237590334757292e-05, "loss": 0.0122, "step": 65450 }, { "grad_norm": 0.2176423817873001, "learning_rate": 2.922254972043543e-05, "loss": 0.0163, "step": 65460 }, { "grad_norm": 0.2533807158470154, "learning_rate": 2.9207511378304265e-05, "loss": 0.0129, "step": 65470 }, { "grad_norm": 0.15002164244651794, "learning_rate": 2.9192475310008428e-05, "loss": 0.0248, "step": 65480 }, { "grad_norm": 0.20256821811199188, "learning_rate": 2.917744151719219e-05, "loss": 0.0137, "step": 65490 }, { "grad_norm": 0.27363288402557373, "learning_rate": 2.916241000149964e-05, "loss": 0.0242, "step": 65500 }, { "grad_norm": 0.21756602823734283, "learning_rate": 2.9147380764574622e-05, "loss": 0.0137, "step": 65510 }, { "grad_norm": 0.052748631685972214, "learning_rate": 2.9132353808060674e-05, "loss": 0.0164, "step": 65520 }, { "grad_norm": 0.10594858974218369, "learning_rate": 2.9117329133601134e-05, "loss": 0.0219, "step": 65530 }, { "grad_norm": 0.19413220882415771, "learning_rate": 2.9102306742839093e-05, "loss": 0.0165, "step": 65540 }, { "grad_norm": 0.17805850505828857, "learning_rate": 2.9087286637417375e-05, "loss": 0.015, "step": 65550 }, { "grad_norm": 0.5500997304916382, "learning_rate": 2.9072268818978537e-05, "loss": 0.0228, "step": 65560 }, { "grad_norm": 0.10875383019447327, "learning_rate": 2.905725328916493e-05, "loss": 0.0142, "step": 65570 }, { "grad_norm": 0.16065001487731934, "learning_rate": 2.9042240049618608e-05, "loss": 0.0251, "step": 65580 }, { "grad_norm": 0.1979164332151413, "learning_rate": 2.9027229101981413e-05, "loss": 0.0157, "step": 65590 }, { "grad_norm": 0.16612043976783752, "learning_rate": 2.901222044789488e-05, "loss": 0.015, "step": 65600 }, { "grad_norm": 0.12528477609157562, "learning_rate": 2.89972140890004e-05, "loss": 0.0134, "step": 65610 }, { "grad_norm": 0.3381485342979431, "learning_rate": 2.8982210026938988e-05, "loss": 0.0149, "step": 65620 }, { "grad_norm": 0.22704215347766876, "learning_rate": 2.896720826335151e-05, "loss": 0.0237, "step": 65630 }, { "grad_norm": 0.37416166067123413, "learning_rate": 2.8952208799878493e-05, "loss": 0.0112, "step": 65640 }, { "grad_norm": 0.40937545895576477, "learning_rate": 2.89372116381603e-05, "loss": 0.0133, "step": 65650 }, { "grad_norm": 0.19254854321479797, "learning_rate": 2.8922216779836932e-05, "loss": 0.012, "step": 65660 }, { "grad_norm": 0.22201229631900787, "learning_rate": 2.8907224226548285e-05, "loss": 0.0195, "step": 65670 }, { "grad_norm": 0.3027280867099762, "learning_rate": 2.8892233979933868e-05, "loss": 0.0187, "step": 65680 }, { "grad_norm": 0.09974700212478638, "learning_rate": 2.8877246041633026e-05, "loss": 0.0194, "step": 65690 }, { "grad_norm": 0.09100082516670227, "learning_rate": 2.886226041328478e-05, "loss": 0.0115, "step": 65700 }, { "grad_norm": 0.15525130927562714, "learning_rate": 2.8847277096527957e-05, "loss": 0.0177, "step": 65710 }, { "grad_norm": 0.22154377400875092, "learning_rate": 2.8832296093001105e-05, "loss": 0.0135, "step": 65720 }, { "grad_norm": 0.4269026815891266, "learning_rate": 2.8817317404342547e-05, "loss": 0.0117, "step": 65730 }, { "grad_norm": 0.10155513882637024, "learning_rate": 2.880234103219029e-05, "loss": 0.0097, "step": 65740 }, { "grad_norm": 0.3066233694553375, "learning_rate": 2.878736697818215e-05, "loss": 0.0129, "step": 65750 }, { "grad_norm": 0.1441659778356552, "learning_rate": 2.877239524395568e-05, "loss": 0.0196, "step": 65760 }, { "grad_norm": 0.10783544927835464, "learning_rate": 2.8757425831148134e-05, "loss": 0.0119, "step": 65770 }, { "grad_norm": 0.284423291683197, "learning_rate": 2.874245874139658e-05, "loss": 0.0173, "step": 65780 }, { "grad_norm": 0.15144003927707672, "learning_rate": 2.8727493976337737e-05, "loss": 0.01, "step": 65790 }, { "grad_norm": 0.18146002292633057, "learning_rate": 2.8712531537608218e-05, "loss": 0.0119, "step": 65800 }, { "grad_norm": 0.1523715704679489, "learning_rate": 2.8697571426844217e-05, "loss": 0.0126, "step": 65810 }, { "grad_norm": 0.08299313485622406, "learning_rate": 2.86826136456818e-05, "loss": 0.0115, "step": 65820 }, { "grad_norm": 0.2143467664718628, "learning_rate": 2.8667658195756685e-05, "loss": 0.0126, "step": 65830 }, { "grad_norm": 0.14613187313079834, "learning_rate": 2.8652705078704412e-05, "loss": 0.0234, "step": 65840 }, { "grad_norm": 0.08522418141365051, "learning_rate": 2.863775429616018e-05, "loss": 0.0133, "step": 65850 }, { "grad_norm": 0.1534961760044098, "learning_rate": 2.862280584975906e-05, "loss": 0.0069, "step": 65860 }, { "grad_norm": 0.48495835065841675, "learning_rate": 2.8607859741135735e-05, "loss": 0.018, "step": 65870 }, { "grad_norm": 0.224994495511055, "learning_rate": 2.8592915971924723e-05, "loss": 0.0156, "step": 65880 }, { "grad_norm": 0.2785697877407074, "learning_rate": 2.8577974543760212e-05, "loss": 0.0157, "step": 65890 }, { "grad_norm": 0.335907518863678, "learning_rate": 2.8563035458276222e-05, "loss": 0.0163, "step": 65900 }, { "grad_norm": 0.12738364934921265, "learning_rate": 2.8548098717106424e-05, "loss": 0.023, "step": 65910 }, { "grad_norm": 0.12786544859409332, "learning_rate": 2.85331643218843e-05, "loss": 0.0101, "step": 65920 }, { "grad_norm": 0.4511075019836426, "learning_rate": 2.8518232274243046e-05, "loss": 0.0113, "step": 65930 }, { "grad_norm": 0.40846529603004456, "learning_rate": 2.8503302575815638e-05, "loss": 0.024, "step": 65940 }, { "grad_norm": 0.1349453330039978, "learning_rate": 2.8488375228234716e-05, "loss": 0.0161, "step": 65950 }, { "grad_norm": 0.20989057421684265, "learning_rate": 2.8473450233132738e-05, "loss": 0.0162, "step": 65960 }, { "grad_norm": 0.11230450123548508, "learning_rate": 2.8458527592141898e-05, "loss": 0.0132, "step": 65970 }, { "grad_norm": 0.19247467815876007, "learning_rate": 2.8443607306894073e-05, "loss": 0.0183, "step": 65980 }, { "grad_norm": 0.11790906637907028, "learning_rate": 2.8428689379020946e-05, "loss": 0.013, "step": 65990 }, { "grad_norm": 0.10356225818395615, "learning_rate": 2.841377381015391e-05, "loss": 0.013, "step": 66000 }, { "grad_norm": 0.291323184967041, "learning_rate": 2.8398860601924144e-05, "loss": 0.0126, "step": 66010 }, { "grad_norm": 0.12338089197874069, "learning_rate": 2.838394975596248e-05, "loss": 0.01, "step": 66020 }, { "grad_norm": 0.07787802815437317, "learning_rate": 2.83690412738996e-05, "loss": 0.0142, "step": 66030 }, { "grad_norm": 0.2780945599079132, "learning_rate": 2.835413515736579e-05, "loss": 0.0236, "step": 66040 }, { "grad_norm": 0.17014773190021515, "learning_rate": 2.8339231407991267e-05, "loss": 0.016, "step": 66050 }, { "grad_norm": 0.31184640526771545, "learning_rate": 2.832433002740581e-05, "loss": 0.0141, "step": 66060 }, { "grad_norm": 0.11835375428199768, "learning_rate": 2.8309431017239047e-05, "loss": 0.0205, "step": 66070 }, { "grad_norm": 0.19476425647735596, "learning_rate": 2.829453437912028e-05, "loss": 0.0168, "step": 66080 }, { "grad_norm": 0.25295311212539673, "learning_rate": 2.8279640114678617e-05, "loss": 0.0141, "step": 66090 }, { "grad_norm": 0.3086215555667877, "learning_rate": 2.8264748225542836e-05, "loss": 0.0164, "step": 66100 }, { "grad_norm": 0.2392515242099762, "learning_rate": 2.8249858713341516e-05, "loss": 0.0121, "step": 66110 }, { "grad_norm": 0.1401560753583908, "learning_rate": 2.823497157970294e-05, "loss": 0.0172, "step": 66120 }, { "grad_norm": 0.24459093809127808, "learning_rate": 2.8220086826255164e-05, "loss": 0.0158, "step": 66130 }, { "grad_norm": 0.12653158605098724, "learning_rate": 2.8205204454625934e-05, "loss": 0.017, "step": 66140 }, { "grad_norm": 0.1888505220413208, "learning_rate": 2.8190324466442798e-05, "loss": 0.0161, "step": 66150 }, { "grad_norm": 0.10292050987482071, "learning_rate": 2.8175446863332965e-05, "loss": 0.0134, "step": 66160 }, { "grad_norm": 0.3162451982498169, "learning_rate": 2.816057164692345e-05, "loss": 0.0184, "step": 66170 }, { "grad_norm": 0.12347757816314697, "learning_rate": 2.8145698818840992e-05, "loss": 0.0151, "step": 66180 }, { "grad_norm": 0.3292911946773529, "learning_rate": 2.8130828380712065e-05, "loss": 0.016, "step": 66190 }, { "grad_norm": 0.051980160176754, "learning_rate": 2.811596033416285e-05, "loss": 0.017, "step": 66200 }, { "grad_norm": 0.30796098709106445, "learning_rate": 2.8101094680819307e-05, "loss": 0.015, "step": 66210 }, { "grad_norm": 0.13277043402194977, "learning_rate": 2.808623142230715e-05, "loss": 0.0147, "step": 66220 }, { "grad_norm": 0.175001323223114, "learning_rate": 2.8071370560251757e-05, "loss": 0.0109, "step": 66230 }, { "grad_norm": 0.1818804144859314, "learning_rate": 2.805651209627831e-05, "loss": 0.0132, "step": 66240 }, { "grad_norm": 0.19975627958774567, "learning_rate": 2.804165603201171e-05, "loss": 0.015, "step": 66250 }, { "grad_norm": 0.26578131318092346, "learning_rate": 2.802680236907661e-05, "loss": 0.0208, "step": 66260 }, { "grad_norm": 0.07731790840625763, "learning_rate": 2.8011951109097344e-05, "loss": 0.015, "step": 66270 }, { "grad_norm": 0.3345559537410736, "learning_rate": 2.7997102253698072e-05, "loss": 0.0161, "step": 66280 }, { "grad_norm": 0.23877865076065063, "learning_rate": 2.7982255804502588e-05, "loss": 0.0136, "step": 66290 }, { "grad_norm": 0.32681965827941895, "learning_rate": 2.7967411763134516e-05, "loss": 0.0144, "step": 66300 }, { "grad_norm": 0.36903679370880127, "learning_rate": 2.7952570131217166e-05, "loss": 0.0198, "step": 66310 }, { "grad_norm": 0.30791744589805603, "learning_rate": 2.7937730910373618e-05, "loss": 0.0193, "step": 66320 }, { "grad_norm": 0.11599688231945038, "learning_rate": 2.7922894102226628e-05, "loss": 0.0142, "step": 66330 }, { "grad_norm": 0.20746146142482758, "learning_rate": 2.790805970839876e-05, "loss": 0.0147, "step": 66340 }, { "grad_norm": 0.23777446150779724, "learning_rate": 2.789322773051225e-05, "loss": 0.0135, "step": 66350 }, { "grad_norm": 0.08760238438844681, "learning_rate": 2.7878398170189124e-05, "loss": 0.0088, "step": 66360 }, { "grad_norm": 0.12391321361064911, "learning_rate": 2.7863571029051112e-05, "loss": 0.0102, "step": 66370 }, { "grad_norm": 0.12580013275146484, "learning_rate": 2.784874630871971e-05, "loss": 0.0162, "step": 66380 }, { "grad_norm": 0.1942029744386673, "learning_rate": 2.7833924010816086e-05, "loss": 0.0116, "step": 66390 }, { "grad_norm": 0.12435676902532578, "learning_rate": 2.7819104136961226e-05, "loss": 0.0139, "step": 66400 }, { "grad_norm": 0.2801946997642517, "learning_rate": 2.780428668877577e-05, "loss": 0.0182, "step": 66410 }, { "grad_norm": 0.3560239374637604, "learning_rate": 2.7789471667880152e-05, "loss": 0.0129, "step": 66420 }, { "grad_norm": 0.10963218659162521, "learning_rate": 2.7774659075894523e-05, "loss": 0.016, "step": 66430 }, { "grad_norm": 0.28804510831832886, "learning_rate": 2.7759848914438778e-05, "loss": 0.017, "step": 66440 }, { "grad_norm": 0.22020334005355835, "learning_rate": 2.7745041185132496e-05, "loss": 0.0116, "step": 66450 }, { "grad_norm": 0.10534775257110596, "learning_rate": 2.773023588959506e-05, "loss": 0.0175, "step": 66460 }, { "grad_norm": 0.12876583635807037, "learning_rate": 2.7715433029445555e-05, "loss": 0.0189, "step": 66470 }, { "grad_norm": 0.3028167486190796, "learning_rate": 2.7700632606302766e-05, "loss": 0.0173, "step": 66480 }, { "grad_norm": 0.09904307872056961, "learning_rate": 2.768583462178528e-05, "loss": 0.0155, "step": 66490 }, { "grad_norm": 0.2610045075416565, "learning_rate": 2.767103907751136e-05, "loss": 0.0126, "step": 66500 }, { "grad_norm": 0.2961300015449524, "learning_rate": 2.7656245975099065e-05, "loss": 0.0131, "step": 66510 }, { "grad_norm": 0.26223111152648926, "learning_rate": 2.7641455316166093e-05, "loss": 0.0141, "step": 66520 }, { "grad_norm": 0.08460713177919388, "learning_rate": 2.7626667102329973e-05, "loss": 0.0128, "step": 66530 }, { "grad_norm": 0.09611520916223526, "learning_rate": 2.761188133520788e-05, "loss": 0.0106, "step": 66540 }, { "grad_norm": 0.08625888079404831, "learning_rate": 2.7597098016416784e-05, "loss": 0.0134, "step": 66550 }, { "grad_norm": 0.1766442060470581, "learning_rate": 2.7582317147573368e-05, "loss": 0.0137, "step": 66560 }, { "grad_norm": 0.10839209705591202, "learning_rate": 2.7567538730294064e-05, "loss": 0.0111, "step": 66570 }, { "grad_norm": 0.18333931267261505, "learning_rate": 2.7552762766194973e-05, "loss": 0.0144, "step": 66580 }, { "grad_norm": 0.2725346088409424, "learning_rate": 2.753798925689202e-05, "loss": 0.0265, "step": 66590 }, { "grad_norm": 0.34162384271621704, "learning_rate": 2.7523218204000774e-05, "loss": 0.0135, "step": 66600 }, { "grad_norm": 0.23688876628875732, "learning_rate": 2.7508449609136612e-05, "loss": 0.017, "step": 66610 }, { "grad_norm": 0.21076630055904388, "learning_rate": 2.749368347391454e-05, "loss": 0.0125, "step": 66620 }, { "grad_norm": 0.33274605870246887, "learning_rate": 2.7478919799949443e-05, "loss": 0.0154, "step": 66630 }, { "grad_norm": 0.22630006074905396, "learning_rate": 2.7464158588855803e-05, "loss": 0.009, "step": 66640 }, { "grad_norm": 0.1419004499912262, "learning_rate": 2.7449399842247918e-05, "loss": 0.0121, "step": 66650 }, { "grad_norm": 0.2612020969390869, "learning_rate": 2.7434643561739736e-05, "loss": 0.0142, "step": 66660 }, { "grad_norm": 0.18266358971595764, "learning_rate": 2.7419889748945016e-05, "loss": 0.0117, "step": 66670 }, { "grad_norm": 0.2277451604604721, "learning_rate": 2.7405138405477203e-05, "loss": 0.0176, "step": 66680 }, { "grad_norm": 0.2725425660610199, "learning_rate": 2.73903895329495e-05, "loss": 0.0107, "step": 66690 }, { "grad_norm": 0.3136318624019623, "learning_rate": 2.737564313297478e-05, "loss": 0.0196, "step": 66700 }, { "grad_norm": 0.1703118085861206, "learning_rate": 2.736089920716572e-05, "loss": 0.0124, "step": 66710 }, { "grad_norm": 0.1352875530719757, "learning_rate": 2.7346157757134706e-05, "loss": 0.0109, "step": 66720 }, { "grad_norm": 0.18218764662742615, "learning_rate": 2.733141878449379e-05, "loss": 0.018, "step": 66730 }, { "grad_norm": 0.7571095824241638, "learning_rate": 2.7316682290854844e-05, "loss": 0.0136, "step": 66740 }, { "grad_norm": 0.16236911714076996, "learning_rate": 2.7301948277829414e-05, "loss": 0.0158, "step": 66750 }, { "grad_norm": 0.3266719579696655, "learning_rate": 2.7287216747028814e-05, "loss": 0.0144, "step": 66760 }, { "grad_norm": 0.11111126840114594, "learning_rate": 2.7272487700064025e-05, "loss": 0.0122, "step": 66770 }, { "grad_norm": 0.29150038957595825, "learning_rate": 2.725776113854583e-05, "loss": 0.0166, "step": 66780 }, { "grad_norm": 0.3991134464740753, "learning_rate": 2.7243037064084664e-05, "loss": 0.0149, "step": 66790 }, { "grad_norm": 0.14389066398143768, "learning_rate": 2.7228315478290766e-05, "loss": 0.0297, "step": 66800 }, { "grad_norm": 0.2023623138666153, "learning_rate": 2.721359638277402e-05, "loss": 0.0147, "step": 66810 }, { "grad_norm": 0.11489767581224442, "learning_rate": 2.7198879779144148e-05, "loss": 0.0095, "step": 66820 }, { "grad_norm": 0.16310156881809235, "learning_rate": 2.7184165669010485e-05, "loss": 0.013, "step": 66830 }, { "grad_norm": 0.14314284920692444, "learning_rate": 2.716945405398218e-05, "loss": 0.0091, "step": 66840 }, { "grad_norm": 0.14014631509780884, "learning_rate": 2.715474493566803e-05, "loss": 0.0128, "step": 66850 }, { "grad_norm": 0.10266433656215668, "learning_rate": 2.7140038315676652e-05, "loss": 0.0137, "step": 66860 }, { "grad_norm": 0.09564507752656937, "learning_rate": 2.7125334195616275e-05, "loss": 0.0428, "step": 66870 }, { "grad_norm": 0.1186705157160759, "learning_rate": 2.7110632577094997e-05, "loss": 0.0165, "step": 66880 }, { "grad_norm": 0.35842859745025635, "learning_rate": 2.709593346172051e-05, "loss": 0.0204, "step": 66890 }, { "grad_norm": 0.17263497412204742, "learning_rate": 2.708123685110032e-05, "loss": 0.013, "step": 66900 }, { "grad_norm": 0.18346676230430603, "learning_rate": 2.7066542746841588e-05, "loss": 0.0119, "step": 66910 }, { "grad_norm": 0.13654282689094543, "learning_rate": 2.7051851150551267e-05, "loss": 0.0118, "step": 66920 }, { "grad_norm": 0.1541842818260193, "learning_rate": 2.7037162063835996e-05, "loss": 0.0159, "step": 66930 }, { "grad_norm": 0.4369472563266754, "learning_rate": 2.702247548830218e-05, "loss": 0.012, "step": 66940 }, { "grad_norm": 0.12223612517118454, "learning_rate": 2.7007791425555883e-05, "loss": 0.0104, "step": 66950 }, { "grad_norm": 0.25608548521995544, "learning_rate": 2.6993109877202945e-05, "loss": 0.0174, "step": 66960 }, { "grad_norm": 0.1699042171239853, "learning_rate": 2.6978430844848935e-05, "loss": 0.0131, "step": 66970 }, { "grad_norm": 0.08806861937046051, "learning_rate": 2.69637543300991e-05, "loss": 0.0109, "step": 66980 }, { "grad_norm": 0.2986718416213989, "learning_rate": 2.694908033455848e-05, "loss": 0.0116, "step": 66990 }, { "grad_norm": 0.07047055661678314, "learning_rate": 2.6934408859831733e-05, "loss": 0.0122, "step": 67000 }, { "grad_norm": 0.18903744220733643, "learning_rate": 2.69197399075234e-05, "loss": 0.0133, "step": 67010 }, { "grad_norm": 0.1643366813659668, "learning_rate": 2.6905073479237584e-05, "loss": 0.0146, "step": 67020 }, { "grad_norm": 0.3632076680660248, "learning_rate": 2.6890409576578235e-05, "loss": 0.0241, "step": 67030 }, { "grad_norm": 0.07539854198694229, "learning_rate": 2.687574820114892e-05, "loss": 0.0151, "step": 67040 }, { "grad_norm": 0.41008782386779785, "learning_rate": 2.686108935455305e-05, "loss": 0.0254, "step": 67050 }, { "grad_norm": 0.549452006816864, "learning_rate": 2.6846433038393603e-05, "loss": 0.0185, "step": 67060 }, { "grad_norm": 0.11174903810024261, "learning_rate": 2.6831779254273482e-05, "loss": 0.0131, "step": 67070 }, { "grad_norm": 0.10167361795902252, "learning_rate": 2.6817128003795122e-05, "loss": 0.0148, "step": 67080 }, { "grad_norm": 0.13367795944213867, "learning_rate": 2.680247928856081e-05, "loss": 0.012, "step": 67090 }, { "grad_norm": 0.14069479703903198, "learning_rate": 2.6787833110172466e-05, "loss": 0.0175, "step": 67100 }, { "grad_norm": 0.05025433003902435, "learning_rate": 2.6773189470231807e-05, "loss": 0.0117, "step": 67110 }, { "grad_norm": 0.23853851854801178, "learning_rate": 2.675854837034022e-05, "loss": 0.0182, "step": 67120 }, { "grad_norm": 0.11988456547260284, "learning_rate": 2.6743909812098833e-05, "loss": 0.0128, "step": 67130 }, { "grad_norm": 0.14044789969921112, "learning_rate": 2.67292737971085e-05, "loss": 0.0205, "step": 67140 }, { "grad_norm": 0.1257762312889099, "learning_rate": 2.671464032696982e-05, "loss": 0.0109, "step": 67150 }, { "grad_norm": 0.3120166063308716, "learning_rate": 2.6700009403283055e-05, "loss": 0.0151, "step": 67160 }, { "grad_norm": 0.17488911747932434, "learning_rate": 2.6685381027648214e-05, "loss": 0.0125, "step": 67170 }, { "grad_norm": 0.24422775208950043, "learning_rate": 2.6670755201665086e-05, "loss": 0.0163, "step": 67180 }, { "grad_norm": 0.3352360725402832, "learning_rate": 2.6656131926933066e-05, "loss": 0.0191, "step": 67190 }, { "grad_norm": 0.22696445882320404, "learning_rate": 2.664151120505136e-05, "loss": 0.01, "step": 67200 }, { "grad_norm": 0.08506464213132858, "learning_rate": 2.6626893037618873e-05, "loss": 0.0161, "step": 67210 }, { "grad_norm": 0.33046573400497437, "learning_rate": 2.6612277426234243e-05, "loss": 0.0159, "step": 67220 }, { "grad_norm": 0.25088995695114136, "learning_rate": 2.659766437249577e-05, "loss": 0.0233, "step": 67230 }, { "grad_norm": 0.11934418231248856, "learning_rate": 2.6583053878001557e-05, "loss": 0.011, "step": 67240 }, { "grad_norm": 0.3619597256183624, "learning_rate": 2.6568445944349323e-05, "loss": 0.0163, "step": 67250 }, { "grad_norm": 0.16890810430049896, "learning_rate": 2.6553840573136657e-05, "loss": 0.0093, "step": 67260 }, { "grad_norm": 0.22412919998168945, "learning_rate": 2.6539237765960706e-05, "loss": 0.0163, "step": 67270 }, { "grad_norm": 0.18101130425930023, "learning_rate": 2.6524637524418465e-05, "loss": 0.0161, "step": 67280 }, { "grad_norm": 0.13340391218662262, "learning_rate": 2.651003985010655e-05, "loss": 0.0117, "step": 67290 }, { "grad_norm": 0.30031999945640564, "learning_rate": 2.649544474462138e-05, "loss": 0.0174, "step": 67300 }, { "grad_norm": 0.1635286957025528, "learning_rate": 2.6480852209559015e-05, "loss": 0.0102, "step": 67310 }, { "grad_norm": 0.27981188893318176, "learning_rate": 2.6466262246515296e-05, "loss": 0.0205, "step": 67320 }, { "grad_norm": 0.19635313749313354, "learning_rate": 2.6451674857085746e-05, "loss": 0.0151, "step": 67330 }, { "grad_norm": 0.24495960772037506, "learning_rate": 2.6437090042865655e-05, "loss": 0.0167, "step": 67340 }, { "grad_norm": 0.3128049671649933, "learning_rate": 2.6422507805449947e-05, "loss": 0.0163, "step": 67350 }, { "grad_norm": 0.277414470911026, "learning_rate": 2.6407928146433357e-05, "loss": 0.0158, "step": 67360 }, { "grad_norm": 0.2142784148454666, "learning_rate": 2.639335106741026e-05, "loss": 0.0105, "step": 67370 }, { "grad_norm": 0.26040858030319214, "learning_rate": 2.6378776569974796e-05, "loss": 0.0213, "step": 67380 }, { "grad_norm": 0.24854780733585358, "learning_rate": 2.6364204655720814e-05, "loss": 0.0137, "step": 67390 }, { "grad_norm": 0.16601231694221497, "learning_rate": 2.6349635326241894e-05, "loss": 0.0203, "step": 67400 }, { "grad_norm": 0.4049220383167267, "learning_rate": 2.633506858313127e-05, "loss": 0.0153, "step": 67410 }, { "grad_norm": 0.3517201244831085, "learning_rate": 2.6320504427981975e-05, "loss": 0.0146, "step": 67420 }, { "grad_norm": 0.13140952587127686, "learning_rate": 2.6305942862386734e-05, "loss": 0.0103, "step": 67430 }, { "grad_norm": 0.1313982456922531, "learning_rate": 2.629138388793793e-05, "loss": 0.0162, "step": 67440 }, { "grad_norm": 0.27100974321365356, "learning_rate": 2.6276827506227743e-05, "loss": 0.0097, "step": 67450 }, { "grad_norm": 0.20358672738075256, "learning_rate": 2.6262273718848037e-05, "loss": 0.0148, "step": 67460 }, { "grad_norm": 0.17509666085243225, "learning_rate": 2.6247722527390407e-05, "loss": 0.0125, "step": 67470 }, { "grad_norm": 0.2621614336967468, "learning_rate": 2.6233173933446104e-05, "loss": 0.0158, "step": 67480 }, { "grad_norm": 0.3139382004737854, "learning_rate": 2.6218627938606188e-05, "loss": 0.025, "step": 67490 }, { "grad_norm": 0.4521685838699341, "learning_rate": 2.620408454446135e-05, "loss": 0.0187, "step": 67500 }, { "grad_norm": 0.12925706803798676, "learning_rate": 2.6189543752602046e-05, "loss": 0.0105, "step": 67510 }, { "grad_norm": 0.2750575840473175, "learning_rate": 2.6175005564618444e-05, "loss": 0.0144, "step": 67520 }, { "grad_norm": 0.2281072437763214, "learning_rate": 2.6160469982100428e-05, "loss": 0.0186, "step": 67530 }, { "grad_norm": 0.3872596025466919, "learning_rate": 2.6145937006637555e-05, "loss": 0.0213, "step": 67540 }, { "grad_norm": 0.20879586040973663, "learning_rate": 2.6131406639819166e-05, "loss": 0.0162, "step": 67550 }, { "grad_norm": 0.4052000343799591, "learning_rate": 2.611687888323424e-05, "loss": 0.0179, "step": 67560 }, { "grad_norm": 0.3126521706581116, "learning_rate": 2.610235373847153e-05, "loss": 0.0194, "step": 67570 }, { "grad_norm": 0.2944582998752594, "learning_rate": 2.6087831207119494e-05, "loss": 0.0182, "step": 67580 }, { "grad_norm": 0.2159016877412796, "learning_rate": 2.60733112907663e-05, "loss": 0.0142, "step": 67590 }, { "grad_norm": 0.1784966140985489, "learning_rate": 2.6058793990999786e-05, "loss": 0.0258, "step": 67600 }, { "grad_norm": 0.35159093141555786, "learning_rate": 2.6044279309407593e-05, "loss": 0.0146, "step": 67610 }, { "grad_norm": 0.18394805490970612, "learning_rate": 2.6029767247576976e-05, "loss": 0.0136, "step": 67620 }, { "grad_norm": 0.11306647211313248, "learning_rate": 2.601525780709497e-05, "loss": 0.0089, "step": 67630 }, { "grad_norm": 0.16070541739463806, "learning_rate": 2.6000750989548318e-05, "loss": 0.0158, "step": 67640 }, { "grad_norm": 0.16233202815055847, "learning_rate": 2.5986246796523462e-05, "loss": 0.011, "step": 67650 }, { "grad_norm": 0.23022863268852234, "learning_rate": 2.5971745229606546e-05, "loss": 0.0107, "step": 67660 }, { "grad_norm": 0.4592929184436798, "learning_rate": 2.5957246290383442e-05, "loss": 0.0179, "step": 67670 }, { "grad_norm": 0.11329983174800873, "learning_rate": 2.594274998043975e-05, "loss": 0.0182, "step": 67680 }, { "grad_norm": 0.1274888515472412, "learning_rate": 2.5928256301360733e-05, "loss": 0.0171, "step": 67690 }, { "grad_norm": 0.12218770384788513, "learning_rate": 2.591376525473142e-05, "loss": 0.0093, "step": 67700 }, { "grad_norm": 0.21582075953483582, "learning_rate": 2.5899276842136523e-05, "loss": 0.0112, "step": 67710 }, { "grad_norm": 0.1967155635356903, "learning_rate": 2.5884791065160495e-05, "loss": 0.0166, "step": 67720 }, { "grad_norm": 0.0963331013917923, "learning_rate": 2.5870307925387436e-05, "loss": 0.0195, "step": 67730 }, { "grad_norm": 0.4749041795730591, "learning_rate": 2.585582742440125e-05, "loss": 0.01, "step": 67740 }, { "grad_norm": 0.17408590018749237, "learning_rate": 2.5841349563785455e-05, "loss": 0.0159, "step": 67750 }, { "grad_norm": 0.25164544582366943, "learning_rate": 2.5826874345123355e-05, "loss": 0.0151, "step": 67760 }, { "grad_norm": 0.1300906240940094, "learning_rate": 2.581240176999792e-05, "loss": 0.0132, "step": 67770 }, { "grad_norm": 0.13788661360740662, "learning_rate": 2.579793183999189e-05, "loss": 0.0092, "step": 67780 }, { "grad_norm": 0.15810880064964294, "learning_rate": 2.5783464556687627e-05, "loss": 0.0212, "step": 67790 }, { "grad_norm": 0.13285823166370392, "learning_rate": 2.576899992166729e-05, "loss": 0.0157, "step": 67800 }, { "grad_norm": 0.20436066389083862, "learning_rate": 2.5754537936512668e-05, "loss": 0.0215, "step": 67810 }, { "grad_norm": 0.29800525307655334, "learning_rate": 2.574007860280535e-05, "loss": 0.013, "step": 67820 }, { "grad_norm": 0.3647686839103699, "learning_rate": 2.5725621922126518e-05, "loss": 0.0186, "step": 67830 }, { "grad_norm": 0.16615460813045502, "learning_rate": 2.571116789605722e-05, "loss": 0.0098, "step": 67840 }, { "grad_norm": 0.47366097569465637, "learning_rate": 2.5696716526178054e-05, "loss": 0.0264, "step": 67850 }, { "grad_norm": 0.17422223091125488, "learning_rate": 2.568226781406946e-05, "loss": 0.0141, "step": 67860 }, { "grad_norm": 0.2417178452014923, "learning_rate": 2.5667821761311474e-05, "loss": 0.0177, "step": 67870 }, { "grad_norm": 0.21680928766727448, "learning_rate": 2.5653378369483917e-05, "loss": 0.0133, "step": 67880 }, { "grad_norm": 0.107977956533432, "learning_rate": 2.5638937640166292e-05, "loss": 0.0136, "step": 67890 }, { "grad_norm": 0.3422999978065491, "learning_rate": 2.5624499574937842e-05, "loss": 0.0137, "step": 67900 }, { "grad_norm": 0.361153244972229, "learning_rate": 2.5610064175377456e-05, "loss": 0.017, "step": 67910 }, { "grad_norm": 0.28653445839881897, "learning_rate": 2.5595631443063782e-05, "loss": 0.0139, "step": 67920 }, { "grad_norm": 0.08176974207162857, "learning_rate": 2.5581201379575177e-05, "loss": 0.0148, "step": 67930 }, { "grad_norm": 0.13563688099384308, "learning_rate": 2.5566773986489655e-05, "loss": 0.0141, "step": 67940 }, { "grad_norm": 0.17592163383960724, "learning_rate": 2.5552349265385e-05, "loss": 0.021, "step": 67950 }, { "grad_norm": 0.335190087556839, "learning_rate": 2.5537927217838675e-05, "loss": 0.0141, "step": 67960 }, { "grad_norm": 0.12384071946144104, "learning_rate": 2.5523507845427864e-05, "loss": 0.0111, "step": 67970 }, { "grad_norm": 0.11505739390850067, "learning_rate": 2.5509091149729413e-05, "loss": 0.0134, "step": 67980 }, { "grad_norm": 0.213857963681221, "learning_rate": 2.5494677132319962e-05, "loss": 0.0202, "step": 67990 }, { "grad_norm": 0.2836472690105438, "learning_rate": 2.548026579477575e-05, "loss": 0.0218, "step": 68000 }, { "grad_norm": 0.26426824927330017, "learning_rate": 2.5465857138672834e-05, "loss": 0.0118, "step": 68010 }, { "grad_norm": 0.19984854757785797, "learning_rate": 2.5451451165586848e-05, "loss": 0.0201, "step": 68020 }, { "grad_norm": 0.20689406991004944, "learning_rate": 2.543704787709329e-05, "loss": 0.0101, "step": 68030 }, { "grad_norm": 0.18564875423908234, "learning_rate": 2.542264727476723e-05, "loss": 0.016, "step": 68040 }, { "grad_norm": 0.14177685976028442, "learning_rate": 2.5408249360183535e-05, "loss": 0.0135, "step": 68050 }, { "grad_norm": 0.17392832040786743, "learning_rate": 2.5393854134916694e-05, "loss": 0.0108, "step": 68060 }, { "grad_norm": 0.11710613965988159, "learning_rate": 2.5379461600540993e-05, "loss": 0.014, "step": 68070 }, { "grad_norm": 0.1560409814119339, "learning_rate": 2.5365071758630317e-05, "loss": 0.0118, "step": 68080 }, { "grad_norm": 0.26002612709999084, "learning_rate": 2.535068461075839e-05, "loss": 0.0156, "step": 68090 }, { "grad_norm": 0.17531079053878784, "learning_rate": 2.5336300158498516e-05, "loss": 0.0118, "step": 68100 }, { "grad_norm": 0.2595042586326599, "learning_rate": 2.5321918403423793e-05, "loss": 0.0122, "step": 68110 }, { "grad_norm": 0.25279518961906433, "learning_rate": 2.5307539347106957e-05, "loss": 0.0173, "step": 68120 }, { "grad_norm": 0.24973709881305695, "learning_rate": 2.529316299112049e-05, "loss": 0.0134, "step": 68130 }, { "grad_norm": 0.1832180768251419, "learning_rate": 2.5278789337036567e-05, "loss": 0.0127, "step": 68140 }, { "grad_norm": 0.3713480532169342, "learning_rate": 2.52644183864271e-05, "loss": 0.0129, "step": 68150 }, { "grad_norm": 0.25517183542251587, "learning_rate": 2.525005014086363e-05, "loss": 0.0149, "step": 68160 }, { "grad_norm": 0.14363497495651245, "learning_rate": 2.5235684601917465e-05, "loss": 0.019, "step": 68170 }, { "grad_norm": 0.1260894536972046, "learning_rate": 2.5221321771159618e-05, "loss": 0.0092, "step": 68180 }, { "grad_norm": 0.16806218028068542, "learning_rate": 2.5206961650160743e-05, "loss": 0.0164, "step": 68190 }, { "grad_norm": 0.18063274025917053, "learning_rate": 2.519260424049129e-05, "loss": 0.0192, "step": 68200 }, { "grad_norm": 0.27702513337135315, "learning_rate": 2.5178249543721304e-05, "loss": 0.0192, "step": 68210 }, { "grad_norm": 0.16168874502182007, "learning_rate": 2.5163897561420668e-05, "loss": 0.0117, "step": 68220 }, { "grad_norm": 0.1417773813009262, "learning_rate": 2.514954829515883e-05, "loss": 0.0165, "step": 68230 }, { "grad_norm": 0.08360502123832703, "learning_rate": 2.513520174650505e-05, "loss": 0.008, "step": 68240 }, { "grad_norm": 0.09684514999389648, "learning_rate": 2.5120857917028207e-05, "loss": 0.0118, "step": 68250 }, { "grad_norm": 0.07313507050275803, "learning_rate": 2.5106516808296953e-05, "loss": 0.0111, "step": 68260 }, { "grad_norm": 0.2520163059234619, "learning_rate": 2.5092178421879552e-05, "loss": 0.0128, "step": 68270 }, { "grad_norm": 0.38744038343429565, "learning_rate": 2.5077842759344116e-05, "loss": 0.0179, "step": 68280 }, { "grad_norm": 0.3099745810031891, "learning_rate": 2.5063509822258304e-05, "loss": 0.0107, "step": 68290 }, { "grad_norm": 0.4112481474876404, "learning_rate": 2.5049179612189576e-05, "loss": 0.0135, "step": 68300 }, { "grad_norm": 0.10613619536161423, "learning_rate": 2.5034852130705043e-05, "loss": 0.0173, "step": 68310 }, { "grad_norm": 0.1459985375404358, "learning_rate": 2.5020527379371572e-05, "loss": 0.0151, "step": 68320 }, { "grad_norm": 0.047993142157793045, "learning_rate": 2.500620535975564e-05, "loss": 0.0081, "step": 68330 }, { "grad_norm": 0.21597760915756226, "learning_rate": 2.4991886073423515e-05, "loss": 0.01, "step": 68340 }, { "grad_norm": 0.17910601198673248, "learning_rate": 2.4977569521941134e-05, "loss": 0.0122, "step": 68350 }, { "grad_norm": 0.203108549118042, "learning_rate": 2.4963255706874143e-05, "loss": 0.0126, "step": 68360 }, { "grad_norm": 0.24923086166381836, "learning_rate": 2.4948944629787853e-05, "loss": 0.0229, "step": 68370 }, { "grad_norm": 0.210082545876503, "learning_rate": 2.493463629224731e-05, "loss": 0.0207, "step": 68380 }, { "grad_norm": 0.13683073222637177, "learning_rate": 2.492033069581728e-05, "loss": 0.0107, "step": 68390 }, { "grad_norm": 0.16512268781661987, "learning_rate": 2.4906027842062137e-05, "loss": 0.0115, "step": 68400 }, { "grad_norm": 0.22556371986865997, "learning_rate": 2.48917277325461e-05, "loss": 0.0087, "step": 68410 }, { "grad_norm": 0.12772250175476074, "learning_rate": 2.487743036883295e-05, "loss": 0.0167, "step": 68420 }, { "grad_norm": 0.15597419440746307, "learning_rate": 2.486313575248626e-05, "loss": 0.01, "step": 68430 }, { "grad_norm": 0.2595323622226715, "learning_rate": 2.4848843885069235e-05, "loss": 0.0144, "step": 68440 }, { "grad_norm": 0.23429147899150848, "learning_rate": 2.483455476814484e-05, "loss": 0.0126, "step": 68450 }, { "grad_norm": 0.09137295931577682, "learning_rate": 2.4820268403275672e-05, "loss": 0.0164, "step": 68460 }, { "grad_norm": 0.3850948214530945, "learning_rate": 2.480598479202413e-05, "loss": 0.0153, "step": 68470 }, { "grad_norm": 0.17572911083698273, "learning_rate": 2.4791703935952193e-05, "loss": 0.0152, "step": 68480 }, { "grad_norm": 0.15770618617534637, "learning_rate": 2.477742583662163e-05, "loss": 0.012, "step": 68490 }, { "grad_norm": 0.29932543635368347, "learning_rate": 2.476315049559384e-05, "loss": 0.0195, "step": 68500 }, { "grad_norm": 0.0896182581782341, "learning_rate": 2.4748877914429987e-05, "loss": 0.025, "step": 68510 }, { "grad_norm": 0.1504071056842804, "learning_rate": 2.4734608094690857e-05, "loss": 0.0172, "step": 68520 }, { "grad_norm": 0.13815920054912567, "learning_rate": 2.4720341037937006e-05, "loss": 0.0179, "step": 68530 }, { "grad_norm": 0.09239044785499573, "learning_rate": 2.4706076745728645e-05, "loss": 0.0247, "step": 68540 }, { "grad_norm": 0.08450745046138763, "learning_rate": 2.469181521962573e-05, "loss": 0.0164, "step": 68550 }, { "grad_norm": 0.26312121748924255, "learning_rate": 2.467755646118783e-05, "loss": 0.007, "step": 68560 }, { "grad_norm": 0.14641070365905762, "learning_rate": 2.46633004719743e-05, "loss": 0.018, "step": 68570 }, { "grad_norm": 0.27561143040657043, "learning_rate": 2.4649047253544123e-05, "loss": 0.0143, "step": 68580 }, { "grad_norm": 0.16682487726211548, "learning_rate": 2.4634796807456016e-05, "loss": 0.0188, "step": 68590 }, { "grad_norm": 0.24709324538707733, "learning_rate": 2.46205491352684e-05, "loss": 0.0152, "step": 68600 }, { "grad_norm": 0.20511335134506226, "learning_rate": 2.4606304238539375e-05, "loss": 0.0169, "step": 68610 }, { "grad_norm": 0.27786359190940857, "learning_rate": 2.4592062118826753e-05, "loss": 0.0123, "step": 68620 }, { "grad_norm": 0.10328016430139542, "learning_rate": 2.4577822777688004e-05, "loss": 0.0117, "step": 68630 }, { "grad_norm": 0.05136710777878761, "learning_rate": 2.4563586216680347e-05, "loss": 0.015, "step": 68640 }, { "grad_norm": 0.15499337017536163, "learning_rate": 2.4549352437360622e-05, "loss": 0.0144, "step": 68650 }, { "grad_norm": 0.13300926983356476, "learning_rate": 2.4535121441285493e-05, "loss": 0.0128, "step": 68660 }, { "grad_norm": 0.2956874370574951, "learning_rate": 2.4520893230011172e-05, "loss": 0.0182, "step": 68670 }, { "grad_norm": 0.12204508483409882, "learning_rate": 2.450666780509368e-05, "loss": 0.0238, "step": 68680 }, { "grad_norm": 0.31459739804267883, "learning_rate": 2.4492445168088656e-05, "loss": 0.0139, "step": 68690 }, { "grad_norm": 0.13822448253631592, "learning_rate": 2.447822532055149e-05, "loss": 0.0087, "step": 68700 }, { "grad_norm": 0.1453849822282791, "learning_rate": 2.4464008264037212e-05, "loss": 0.0206, "step": 68710 }, { "grad_norm": 0.21766260266304016, "learning_rate": 2.4449794000100605e-05, "loss": 0.0126, "step": 68720 }, { "grad_norm": 0.06527957320213318, "learning_rate": 2.4435582530296108e-05, "loss": 0.0192, "step": 68730 }, { "grad_norm": 0.11328040808439255, "learning_rate": 2.4421373856177887e-05, "loss": 0.0125, "step": 68740 }, { "grad_norm": 0.2413562387228012, "learning_rate": 2.440716797929975e-05, "loss": 0.0127, "step": 68750 }, { "grad_norm": 0.11441808938980103, "learning_rate": 2.439296490121526e-05, "loss": 0.0121, "step": 68760 }, { "grad_norm": 0.10672245174646378, "learning_rate": 2.4378764623477608e-05, "loss": 0.0078, "step": 68770 }, { "grad_norm": 0.24151615798473358, "learning_rate": 2.436456714763974e-05, "loss": 0.0143, "step": 68780 }, { "grad_norm": 0.14001469314098358, "learning_rate": 2.4350372475254262e-05, "loss": 0.0185, "step": 68790 }, { "grad_norm": 0.14521200954914093, "learning_rate": 2.433618060787351e-05, "loss": 0.0131, "step": 68800 }, { "grad_norm": 0.07943248003721237, "learning_rate": 2.432199154704944e-05, "loss": 0.0225, "step": 68810 }, { "grad_norm": 0.22060339152812958, "learning_rate": 2.430780529433377e-05, "loss": 0.0116, "step": 68820 }, { "grad_norm": 0.09041287004947662, "learning_rate": 2.42936218512779e-05, "loss": 0.0242, "step": 68830 }, { "grad_norm": 0.17962904274463654, "learning_rate": 2.427944121943288e-05, "loss": 0.0085, "step": 68840 }, { "grad_norm": 0.11401179432868958, "learning_rate": 2.4265263400349497e-05, "loss": 0.0114, "step": 68850 }, { "grad_norm": 0.5155532956123352, "learning_rate": 2.4251088395578213e-05, "loss": 0.0095, "step": 68860 }, { "grad_norm": 0.5546783208847046, "learning_rate": 2.4236916206669214e-05, "loss": 0.0138, "step": 68870 }, { "grad_norm": 0.11102909594774246, "learning_rate": 2.4222746835172304e-05, "loss": 0.0103, "step": 68880 }, { "grad_norm": 0.4529905319213867, "learning_rate": 2.4208580282637065e-05, "loss": 0.0208, "step": 68890 }, { "grad_norm": 0.29792672395706177, "learning_rate": 2.4194416550612687e-05, "loss": 0.014, "step": 68900 }, { "grad_norm": 0.205275759100914, "learning_rate": 2.418025564064812e-05, "loss": 0.0122, "step": 68910 }, { "grad_norm": 0.08407266438007355, "learning_rate": 2.4166097554291978e-05, "loss": 0.0118, "step": 68920 }, { "grad_norm": 0.10610266774892807, "learning_rate": 2.4151942293092583e-05, "loss": 0.0115, "step": 68930 }, { "grad_norm": 0.24767999351024628, "learning_rate": 2.41377898585979e-05, "loss": 0.0136, "step": 68940 }, { "grad_norm": 0.4568157494068146, "learning_rate": 2.4123640252355652e-05, "loss": 0.0186, "step": 68950 }, { "grad_norm": 0.1845189929008484, "learning_rate": 2.410949347591318e-05, "loss": 0.009, "step": 68960 }, { "grad_norm": 0.3199548125267029, "learning_rate": 2.4095349530817578e-05, "loss": 0.0283, "step": 68970 }, { "grad_norm": 0.3487050533294678, "learning_rate": 2.4081208418615603e-05, "loss": 0.0152, "step": 68980 }, { "grad_norm": 0.23436471819877625, "learning_rate": 2.4067070140853725e-05, "loss": 0.0115, "step": 68990 }, { "grad_norm": 0.24387578666210175, "learning_rate": 2.405293469907805e-05, "loss": 0.0136, "step": 69000 }, { "grad_norm": 0.31158632040023804, "learning_rate": 2.4038802094834446e-05, "loss": 0.0137, "step": 69010 }, { "grad_norm": 0.16092145442962646, "learning_rate": 2.4024672329668395e-05, "loss": 0.0139, "step": 69020 }, { "grad_norm": 0.3176150619983673, "learning_rate": 2.4010545405125125e-05, "loss": 0.0117, "step": 69030 }, { "grad_norm": 0.3222760558128357, "learning_rate": 2.3996421322749528e-05, "loss": 0.0209, "step": 69040 }, { "grad_norm": 0.38814812898635864, "learning_rate": 2.3982300084086224e-05, "loss": 0.0191, "step": 69050 }, { "grad_norm": 0.18587058782577515, "learning_rate": 2.396818169067944e-05, "loss": 0.0096, "step": 69060 }, { "grad_norm": 0.07390189915895462, "learning_rate": 2.3954066144073178e-05, "loss": 0.0121, "step": 69070 }, { "grad_norm": 0.19397802650928497, "learning_rate": 2.39399534458111e-05, "loss": 0.0213, "step": 69080 }, { "grad_norm": 0.0799574926495552, "learning_rate": 2.392584359743651e-05, "loss": 0.0135, "step": 69090 }, { "grad_norm": 0.16047276556491852, "learning_rate": 2.3911736600492463e-05, "loss": 0.0124, "step": 69100 }, { "grad_norm": 0.43350571393966675, "learning_rate": 2.3897632456521672e-05, "loss": 0.0165, "step": 69110 }, { "grad_norm": 0.21182796359062195, "learning_rate": 2.388353116706658e-05, "loss": 0.0321, "step": 69120 }, { "grad_norm": 0.14104676246643066, "learning_rate": 2.3869432733669232e-05, "loss": 0.0128, "step": 69130 }, { "grad_norm": 0.12712806463241577, "learning_rate": 2.385533715787145e-05, "loss": 0.0137, "step": 69140 }, { "grad_norm": 0.3214115500450134, "learning_rate": 2.3841244441214673e-05, "loss": 0.0143, "step": 69150 }, { "grad_norm": 0.2656717002391815, "learning_rate": 2.3827154585240073e-05, "loss": 0.0196, "step": 69160 }, { "grad_norm": 0.2092311829328537, "learning_rate": 2.3813067591488498e-05, "loss": 0.0186, "step": 69170 }, { "grad_norm": 0.09145979583263397, "learning_rate": 2.3798983461500494e-05, "loss": 0.0172, "step": 69180 }, { "grad_norm": 0.1578817516565323, "learning_rate": 2.3784902196816256e-05, "loss": 0.0146, "step": 69190 }, { "grad_norm": 0.16702768206596375, "learning_rate": 2.3770823798975716e-05, "loss": 0.0165, "step": 69200 }, { "grad_norm": 0.22571338713169098, "learning_rate": 2.3756748269518424e-05, "loss": 0.0123, "step": 69210 }, { "grad_norm": 0.11755915731191635, "learning_rate": 2.374267560998371e-05, "loss": 0.0139, "step": 69220 }, { "grad_norm": 0.19958297908306122, "learning_rate": 2.3728605821910477e-05, "loss": 0.0144, "step": 69230 }, { "grad_norm": 0.30835989117622375, "learning_rate": 2.3714538906837453e-05, "loss": 0.023, "step": 69240 }, { "grad_norm": 0.1562037169933319, "learning_rate": 2.3700474866302913e-05, "loss": 0.012, "step": 69250 }, { "grad_norm": 0.26094046235084534, "learning_rate": 2.3686413701844917e-05, "loss": 0.0108, "step": 69260 }, { "grad_norm": 0.12293649464845657, "learning_rate": 2.367235541500114e-05, "loss": 0.0099, "step": 69270 }, { "grad_norm": 0.08277155458927155, "learning_rate": 2.3658300007308993e-05, "loss": 0.0138, "step": 69280 }, { "grad_norm": 0.365951806306839, "learning_rate": 2.364424748030555e-05, "loss": 0.0089, "step": 69290 }, { "grad_norm": 0.1635051667690277, "learning_rate": 2.3630197835527595e-05, "loss": 0.0167, "step": 69300 }, { "grad_norm": 0.09056557714939117, "learning_rate": 2.3616151074511546e-05, "loss": 0.022, "step": 69310 }, { "grad_norm": 0.13844947516918182, "learning_rate": 2.3602107198793545e-05, "loss": 0.023, "step": 69320 }, { "grad_norm": 0.1101449579000473, "learning_rate": 2.3588066209909427e-05, "loss": 0.0155, "step": 69330 }, { "grad_norm": 0.08527577668428421, "learning_rate": 2.3574028109394662e-05, "loss": 0.0142, "step": 69340 }, { "grad_norm": 0.1326196789741516, "learning_rate": 2.3559992898784454e-05, "loss": 0.0184, "step": 69350 }, { "grad_norm": 0.0843840166926384, "learning_rate": 2.354596057961367e-05, "loss": 0.0144, "step": 69360 }, { "grad_norm": 0.23596970736980438, "learning_rate": 2.353193115341688e-05, "loss": 0.0253, "step": 69370 }, { "grad_norm": 0.16566713154315948, "learning_rate": 2.3517904621728294e-05, "loss": 0.015, "step": 69380 }, { "grad_norm": 0.20825858414173126, "learning_rate": 2.3503880986081855e-05, "loss": 0.0134, "step": 69390 }, { "grad_norm": 0.15050728619098663, "learning_rate": 2.3489860248011136e-05, "loss": 0.0198, "step": 69400 }, { "grad_norm": 0.12514416873455048, "learning_rate": 2.3475842409049464e-05, "loss": 0.0113, "step": 69410 }, { "grad_norm": 0.1839314103126526, "learning_rate": 2.346182747072975e-05, "loss": 0.0196, "step": 69420 }, { "grad_norm": 0.1431528776884079, "learning_rate": 2.3447815434584718e-05, "loss": 0.0119, "step": 69430 }, { "grad_norm": 0.16661712527275085, "learning_rate": 2.3433806302146656e-05, "loss": 0.0101, "step": 69440 }, { "grad_norm": 0.26778656244277954, "learning_rate": 2.3419800074947612e-05, "loss": 0.0188, "step": 69450 }, { "grad_norm": 0.10301098227500916, "learning_rate": 2.3405796754519248e-05, "loss": 0.0175, "step": 69460 }, { "grad_norm": 0.24546338617801666, "learning_rate": 2.3391796342392992e-05, "loss": 0.012, "step": 69470 }, { "grad_norm": 0.34652209281921387, "learning_rate": 2.337779884009984e-05, "loss": 0.0182, "step": 69480 }, { "grad_norm": 0.08305848389863968, "learning_rate": 2.3363804249170624e-05, "loss": 0.0091, "step": 69490 }, { "grad_norm": 0.4523138403892517, "learning_rate": 2.3349812571135703e-05, "loss": 0.0139, "step": 69500 }, { "grad_norm": 0.15003007650375366, "learning_rate": 2.333582380752523e-05, "loss": 0.0146, "step": 69510 }, { "grad_norm": 0.07989946752786636, "learning_rate": 2.3321837959868963e-05, "loss": 0.0108, "step": 69520 }, { "grad_norm": 0.26079630851745605, "learning_rate": 2.3307855029696383e-05, "loss": 0.0109, "step": 69530 }, { "grad_norm": 0.3553195595741272, "learning_rate": 2.3293875018536648e-05, "loss": 0.0118, "step": 69540 }, { "grad_norm": 0.1263095736503601, "learning_rate": 2.327989792791861e-05, "loss": 0.012, "step": 69550 }, { "grad_norm": 0.1506977379322052, "learning_rate": 2.326592375937074e-05, "loss": 0.0134, "step": 69560 }, { "grad_norm": 0.128244087100029, "learning_rate": 2.3251952514421248e-05, "loss": 0.0122, "step": 69570 }, { "grad_norm": 0.26606157422065735, "learning_rate": 2.323798419459804e-05, "loss": 0.0126, "step": 69580 }, { "grad_norm": 0.2325619012117386, "learning_rate": 2.3224018801428625e-05, "loss": 0.0101, "step": 69590 }, { "grad_norm": 0.194054514169693, "learning_rate": 2.3210056336440273e-05, "loss": 0.0142, "step": 69600 }, { "grad_norm": 0.486368328332901, "learning_rate": 2.3196096801159844e-05, "loss": 0.0174, "step": 69610 }, { "grad_norm": 0.23732677102088928, "learning_rate": 2.3182140197114012e-05, "loss": 0.0164, "step": 69620 }, { "grad_norm": 0.25489541888237, "learning_rate": 2.3168186525828985e-05, "loss": 0.0255, "step": 69630 }, { "grad_norm": 0.1278354823589325, "learning_rate": 2.3154235788830757e-05, "loss": 0.0135, "step": 69640 }, { "grad_norm": 0.25305551290512085, "learning_rate": 2.3140287987644927e-05, "loss": 0.0123, "step": 69650 }, { "grad_norm": 0.19945409893989563, "learning_rate": 2.3126343123796833e-05, "loss": 0.0139, "step": 69660 }, { "grad_norm": 0.07470139116048813, "learning_rate": 2.3112401198811418e-05, "loss": 0.0152, "step": 69670 }, { "grad_norm": 0.28846248984336853, "learning_rate": 2.3098462214213418e-05, "loss": 0.0156, "step": 69680 }, { "grad_norm": 0.14529334008693695, "learning_rate": 2.308452617152712e-05, "loss": 0.0171, "step": 69690 }, { "grad_norm": 0.12013374269008636, "learning_rate": 2.3070593072276603e-05, "loss": 0.0132, "step": 69700 }, { "grad_norm": 0.17227140069007874, "learning_rate": 2.3056662917985518e-05, "loss": 0.0116, "step": 69710 }, { "grad_norm": 0.27703240513801575, "learning_rate": 2.3042735710177283e-05, "loss": 0.0297, "step": 69720 }, { "grad_norm": 0.34803441166877747, "learning_rate": 2.302881145037492e-05, "loss": 0.0117, "step": 69730 }, { "grad_norm": 0.3792979419231415, "learning_rate": 2.3014890140101192e-05, "loss": 0.0192, "step": 69740 }, { "grad_norm": 0.15501317381858826, "learning_rate": 2.30009717808785e-05, "loss": 0.0106, "step": 69750 }, { "grad_norm": 0.13671645522117615, "learning_rate": 2.2987056374228967e-05, "loss": 0.0138, "step": 69760 }, { "grad_norm": 0.05411756783723831, "learning_rate": 2.297314392167431e-05, "loss": 0.0119, "step": 69770 }, { "grad_norm": 0.41978174448013306, "learning_rate": 2.295923442473601e-05, "loss": 0.0182, "step": 69780 }, { "grad_norm": 0.8087447285652161, "learning_rate": 2.294532788493519e-05, "loss": 0.0146, "step": 69790 }, { "grad_norm": 0.22953344881534576, "learning_rate": 2.293142430379262e-05, "loss": 0.0134, "step": 69800 }, { "grad_norm": 0.19794988632202148, "learning_rate": 2.291752368282879e-05, "loss": 0.01, "step": 69810 }, { "grad_norm": 0.21906088292598724, "learning_rate": 2.290362602356385e-05, "loss": 0.0146, "step": 69820 }, { "grad_norm": 0.23525558412075043, "learning_rate": 2.288973132751765e-05, "loss": 0.0117, "step": 69830 }, { "grad_norm": 0.18237566947937012, "learning_rate": 2.287583959620965e-05, "loss": 0.0205, "step": 69840 }, { "grad_norm": 0.09469158947467804, "learning_rate": 2.2861950831159074e-05, "loss": 0.0135, "step": 69850 }, { "grad_norm": 0.23775553703308105, "learning_rate": 2.284806503388472e-05, "loss": 0.0168, "step": 69860 }, { "grad_norm": 0.12688776850700378, "learning_rate": 2.283418220590518e-05, "loss": 0.0171, "step": 69870 }, { "grad_norm": 0.19633111357688904, "learning_rate": 2.282030234873862e-05, "loss": 0.0142, "step": 69880 }, { "grad_norm": 0.19158628582954407, "learning_rate": 2.2806425463902948e-05, "loss": 0.0137, "step": 69890 }, { "grad_norm": 0.28460782766342163, "learning_rate": 2.279255155291568e-05, "loss": 0.0215, "step": 69900 }, { "grad_norm": 0.14713667333126068, "learning_rate": 2.277868061729409e-05, "loss": 0.0104, "step": 69910 }, { "grad_norm": 0.12166264653205872, "learning_rate": 2.276481265855504e-05, "loss": 0.0188, "step": 69920 }, { "grad_norm": 0.17435583472251892, "learning_rate": 2.2750947678215128e-05, "loss": 0.0168, "step": 69930 }, { "grad_norm": 0.10410621017217636, "learning_rate": 2.2737085677790616e-05, "loss": 0.0084, "step": 69940 }, { "grad_norm": 0.15875761210918427, "learning_rate": 2.2723226658797443e-05, "loss": 0.0155, "step": 69950 }, { "grad_norm": 0.1910601705312729, "learning_rate": 2.270937062275117e-05, "loss": 0.0206, "step": 69960 }, { "grad_norm": 0.25611424446105957, "learning_rate": 2.2695517571167113e-05, "loss": 0.0161, "step": 69970 }, { "grad_norm": 0.07497396320104599, "learning_rate": 2.268166750556019e-05, "loss": 0.0124, "step": 69980 }, { "grad_norm": 0.1869773268699646, "learning_rate": 2.266782042744503e-05, "loss": 0.0125, "step": 69990 }, { "grad_norm": 0.13560587167739868, "learning_rate": 2.2653976338335935e-05, "loss": 0.0114, "step": 70000 }, { "grad_norm": 0.3838241398334503, "learning_rate": 2.2640135239746895e-05, "loss": 0.0167, "step": 70010 }, { "grad_norm": 0.23032160103321075, "learning_rate": 2.2626297133191515e-05, "loss": 0.0138, "step": 70020 }, { "grad_norm": 0.11043737828731537, "learning_rate": 2.261246202018312e-05, "loss": 0.0108, "step": 70030 }, { "grad_norm": 0.07577262818813324, "learning_rate": 2.2598629902234718e-05, "loss": 0.0075, "step": 70040 }, { "grad_norm": 0.16293078660964966, "learning_rate": 2.258480078085894e-05, "loss": 0.0145, "step": 70050 }, { "grad_norm": 0.1638675332069397, "learning_rate": 2.2570974657568126e-05, "loss": 0.0086, "step": 70060 }, { "grad_norm": 0.10032624751329422, "learning_rate": 2.2557151533874286e-05, "loss": 0.0104, "step": 70070 }, { "grad_norm": 0.3827023208141327, "learning_rate": 2.2543331411289116e-05, "loss": 0.0095, "step": 70080 }, { "grad_norm": 0.18863807618618011, "learning_rate": 2.252951429132392e-05, "loss": 0.0116, "step": 70090 }, { "grad_norm": 0.10467683523893356, "learning_rate": 2.2515700175489762e-05, "loss": 0.0105, "step": 70100 }, { "grad_norm": 0.1449129432439804, "learning_rate": 2.2501889065297288e-05, "loss": 0.0077, "step": 70110 }, { "grad_norm": 0.2852831780910492, "learning_rate": 2.2488080962256887e-05, "loss": 0.0144, "step": 70120 }, { "grad_norm": 0.19437502324581146, "learning_rate": 2.2474275867878585e-05, "loss": 0.0142, "step": 70130 }, { "grad_norm": 0.3196329176425934, "learning_rate": 2.246047378367211e-05, "loss": 0.0219, "step": 70140 }, { "grad_norm": 0.11244713515043259, "learning_rate": 2.2446674711146798e-05, "loss": 0.0131, "step": 70150 }, { "grad_norm": 0.09278757870197296, "learning_rate": 2.2432878651811734e-05, "loss": 0.0169, "step": 70160 }, { "grad_norm": 0.1176055520772934, "learning_rate": 2.2419085607175594e-05, "loss": 0.0155, "step": 70170 }, { "grad_norm": 0.136699378490448, "learning_rate": 2.240529557874678e-05, "loss": 0.0212, "step": 70180 }, { "grad_norm": 0.36661139130592346, "learning_rate": 2.2391508568033358e-05, "loss": 0.0126, "step": 70190 }, { "grad_norm": 0.24402199685573578, "learning_rate": 2.237772457654307e-05, "loss": 0.0113, "step": 70200 }, { "grad_norm": 0.35143396258354187, "learning_rate": 2.236394360578327e-05, "loss": 0.0143, "step": 70210 }, { "grad_norm": 0.2556547224521637, "learning_rate": 2.2350165657261073e-05, "loss": 0.0106, "step": 70220 }, { "grad_norm": 0.18041160702705383, "learning_rate": 2.2336390732483164e-05, "loss": 0.0165, "step": 70230 }, { "grad_norm": 0.22403796017169952, "learning_rate": 2.2322618832955976e-05, "loss": 0.0171, "step": 70240 }, { "grad_norm": 0.14620240032672882, "learning_rate": 2.2308849960185578e-05, "loss": 0.0138, "step": 70250 }, { "grad_norm": 0.09189844876527786, "learning_rate": 2.229508411567774e-05, "loss": 0.0133, "step": 70260 }, { "grad_norm": 0.24081681668758392, "learning_rate": 2.2281321300937823e-05, "loss": 0.0193, "step": 70270 }, { "grad_norm": 0.12453518807888031, "learning_rate": 2.2267561517470937e-05, "loss": 0.0117, "step": 70280 }, { "grad_norm": 0.10248339921236038, "learning_rate": 2.2253804766781845e-05, "loss": 0.0176, "step": 70290 }, { "grad_norm": 0.45606255531311035, "learning_rate": 2.2240051050374933e-05, "loss": 0.0186, "step": 70300 }, { "grad_norm": 0.1700468361377716, "learning_rate": 2.22263003697543e-05, "loss": 0.0148, "step": 70310 }, { "grad_norm": 0.215341255068779, "learning_rate": 2.2212552726423702e-05, "loss": 0.0204, "step": 70320 }, { "grad_norm": 0.10917025059461594, "learning_rate": 2.2198808121886578e-05, "loss": 0.0135, "step": 70330 }, { "grad_norm": 0.12651893496513367, "learning_rate": 2.218506655764598e-05, "loss": 0.0151, "step": 70340 }, { "grad_norm": 0.16357527673244476, "learning_rate": 2.2171328035204704e-05, "loss": 0.0138, "step": 70350 }, { "grad_norm": 0.1251232773065567, "learning_rate": 2.2157592556065133e-05, "loss": 0.0129, "step": 70360 }, { "grad_norm": 0.25397926568984985, "learning_rate": 2.2143860121729382e-05, "loss": 0.0096, "step": 70370 }, { "grad_norm": 0.15944801270961761, "learning_rate": 2.2130130733699206e-05, "loss": 0.01, "step": 70380 }, { "grad_norm": 0.16708700358867645, "learning_rate": 2.2116404393476052e-05, "loss": 0.0131, "step": 70390 }, { "grad_norm": 0.209467813372612, "learning_rate": 2.2102681102560967e-05, "loss": 0.0194, "step": 70400 }, { "grad_norm": 0.17869621515274048, "learning_rate": 2.208896086245476e-05, "loss": 0.0176, "step": 70410 }, { "grad_norm": 0.24640972912311554, "learning_rate": 2.2075243674657808e-05, "loss": 0.0141, "step": 70420 }, { "grad_norm": 0.08160028606653214, "learning_rate": 2.206152954067024e-05, "loss": 0.0076, "step": 70430 }, { "grad_norm": 0.1439819037914276, "learning_rate": 2.2047818461991758e-05, "loss": 0.0228, "step": 70440 }, { "grad_norm": 0.27663299441337585, "learning_rate": 2.2034110440121864e-05, "loss": 0.013, "step": 70450 }, { "grad_norm": 0.1470179706811905, "learning_rate": 2.2020405476559585e-05, "loss": 0.012, "step": 70460 }, { "grad_norm": 0.20878037810325623, "learning_rate": 2.2006703572803722e-05, "loss": 0.0139, "step": 70470 }, { "grad_norm": 0.06017564237117767, "learning_rate": 2.1993004730352646e-05, "loss": 0.0121, "step": 70480 }, { "grad_norm": 0.29255157709121704, "learning_rate": 2.1979308950704464e-05, "loss": 0.0146, "step": 70490 }, { "grad_norm": 0.18419945240020752, "learning_rate": 2.1965616235356923e-05, "loss": 0.0149, "step": 70500 }, { "grad_norm": 0.10096029192209244, "learning_rate": 2.1951926585807465e-05, "loss": 0.0107, "step": 70510 }, { "grad_norm": 0.11080571264028549, "learning_rate": 2.193824000355313e-05, "loss": 0.0093, "step": 70520 }, { "grad_norm": 0.3208547830581665, "learning_rate": 2.192455649009067e-05, "loss": 0.0137, "step": 70530 }, { "grad_norm": 0.12680283188819885, "learning_rate": 2.1910876046916523e-05, "loss": 0.0111, "step": 70540 }, { "grad_norm": 0.27497953176498413, "learning_rate": 2.1897198675526725e-05, "loss": 0.0116, "step": 70550 }, { "grad_norm": 0.17188690602779388, "learning_rate": 2.1883524377417024e-05, "loss": 0.0146, "step": 70560 }, { "grad_norm": 0.19702620804309845, "learning_rate": 2.1869853154082825e-05, "loss": 0.0143, "step": 70570 }, { "grad_norm": 0.24826441705226898, "learning_rate": 2.1856185007019215e-05, "loss": 0.0127, "step": 70580 }, { "grad_norm": 0.169541135430336, "learning_rate": 2.1842519937720874e-05, "loss": 0.012, "step": 70590 }, { "grad_norm": 0.07796737551689148, "learning_rate": 2.182885794768224e-05, "loss": 0.0148, "step": 70600 }, { "grad_norm": 0.3921331465244293, "learning_rate": 2.1815199038397326e-05, "loss": 0.0109, "step": 70610 }, { "grad_norm": 0.24955952167510986, "learning_rate": 2.1801543211359893e-05, "loss": 0.016, "step": 70620 }, { "grad_norm": 0.13262546062469482, "learning_rate": 2.178789046806326e-05, "loss": 0.013, "step": 70630 }, { "grad_norm": 0.2388371378183365, "learning_rate": 2.1774240810000547e-05, "loss": 0.0214, "step": 70640 }, { "grad_norm": 0.11992253363132477, "learning_rate": 2.17605942386644e-05, "loss": 0.0188, "step": 70650 }, { "grad_norm": 0.3332240879535675, "learning_rate": 2.174695075554723e-05, "loss": 0.02, "step": 70660 }, { "grad_norm": 0.25443780422210693, "learning_rate": 2.173331036214103e-05, "loss": 0.0224, "step": 70670 }, { "grad_norm": 0.14460326731204987, "learning_rate": 2.171967305993753e-05, "loss": 0.0132, "step": 70680 }, { "grad_norm": 0.2426152527332306, "learning_rate": 2.1706038850428024e-05, "loss": 0.0164, "step": 70690 }, { "grad_norm": 0.23674815893173218, "learning_rate": 2.169240773510361e-05, "loss": 0.0138, "step": 70700 }, { "grad_norm": 0.08259011805057526, "learning_rate": 2.1678779715454906e-05, "loss": 0.0132, "step": 70710 }, { "grad_norm": 0.09770029038190842, "learning_rate": 2.166515479297229e-05, "loss": 0.0123, "step": 70720 }, { "grad_norm": 0.07120510935783386, "learning_rate": 2.1651532969145732e-05, "loss": 0.0058, "step": 70730 }, { "grad_norm": 0.402175635099411, "learning_rate": 2.1637914245464903e-05, "loss": 0.0121, "step": 70740 }, { "grad_norm": 0.3875875771045685, "learning_rate": 2.1624298623419133e-05, "loss": 0.0131, "step": 70750 }, { "grad_norm": 0.36432328820228577, "learning_rate": 2.1610686104497413e-05, "loss": 0.0137, "step": 70760 }, { "grad_norm": 0.47323378920555115, "learning_rate": 2.1597076690188366e-05, "loss": 0.0146, "step": 70770 }, { "grad_norm": 0.14174848794937134, "learning_rate": 2.1583470381980315e-05, "loss": 0.0098, "step": 70780 }, { "grad_norm": 0.14264892041683197, "learning_rate": 2.156986718136123e-05, "loss": 0.0209, "step": 70790 }, { "grad_norm": 0.3870904743671417, "learning_rate": 2.155626708981871e-05, "loss": 0.0136, "step": 70800 }, { "grad_norm": 0.11502952873706818, "learning_rate": 2.1542670108840075e-05, "loss": 0.0119, "step": 70810 }, { "grad_norm": 0.9889637231826782, "learning_rate": 2.152907623991221e-05, "loss": 0.0177, "step": 70820 }, { "grad_norm": 0.47640952467918396, "learning_rate": 2.1515485484521802e-05, "loss": 0.0143, "step": 70830 }, { "grad_norm": 0.09216365963220596, "learning_rate": 2.150189784415506e-05, "loss": 0.0092, "step": 70840 }, { "grad_norm": 0.1296597421169281, "learning_rate": 2.148831332029794e-05, "loss": 0.0117, "step": 70850 }, { "grad_norm": 0.11284570395946503, "learning_rate": 2.147473191443599e-05, "loss": 0.0159, "step": 70860 }, { "grad_norm": 0.18914245069026947, "learning_rate": 2.1461153628054492e-05, "loss": 0.0136, "step": 70870 }, { "grad_norm": 0.1493191123008728, "learning_rate": 2.1447578462638284e-05, "loss": 0.015, "step": 70880 }, { "grad_norm": 0.1944229155778885, "learning_rate": 2.143400641967201e-05, "loss": 0.0135, "step": 70890 }, { "grad_norm": 0.19093601405620575, "learning_rate": 2.1420437500639823e-05, "loss": 0.0115, "step": 70900 }, { "grad_norm": 0.1599780023097992, "learning_rate": 2.1406871707025646e-05, "loss": 0.012, "step": 70910 }, { "grad_norm": 0.07681140303611755, "learning_rate": 2.1393309040312963e-05, "loss": 0.025, "step": 70920 }, { "grad_norm": 0.2845900058746338, "learning_rate": 2.137974950198502e-05, "loss": 0.0121, "step": 70930 }, { "grad_norm": 0.34704622626304626, "learning_rate": 2.1366193093524617e-05, "loss": 0.0118, "step": 70940 }, { "grad_norm": 0.11107242852449417, "learning_rate": 2.1352639816414288e-05, "loss": 0.0087, "step": 70950 }, { "grad_norm": 0.30083879828453064, "learning_rate": 2.1339089672136202e-05, "loss": 0.0198, "step": 70960 }, { "grad_norm": 0.1073094978928566, "learning_rate": 2.1325542662172193e-05, "loss": 0.0108, "step": 70970 }, { "grad_norm": 0.2339164763689041, "learning_rate": 2.1311998788003712e-05, "loss": 0.013, "step": 70980 }, { "grad_norm": 0.1380470246076584, "learning_rate": 2.1298458051111907e-05, "loss": 0.0091, "step": 70990 }, { "grad_norm": 0.21200287342071533, "learning_rate": 2.1284920452977602e-05, "loss": 0.0142, "step": 71000 }, { "grad_norm": 0.2052682489156723, "learning_rate": 2.12713859950812e-05, "loss": 0.0133, "step": 71010 }, { "grad_norm": 0.2456609457731247, "learning_rate": 2.1257854678902832e-05, "loss": 0.0147, "step": 71020 }, { "grad_norm": 0.17417851090431213, "learning_rate": 2.124432650592227e-05, "loss": 0.0154, "step": 71030 }, { "grad_norm": 0.31106865406036377, "learning_rate": 2.1230801477618937e-05, "loss": 0.0167, "step": 71040 }, { "grad_norm": 0.5680220127105713, "learning_rate": 2.121727959547189e-05, "loss": 0.0213, "step": 71050 }, { "grad_norm": 0.24111641943454742, "learning_rate": 2.1203760860959886e-05, "loss": 0.0203, "step": 71060 }, { "grad_norm": 0.17682676017284393, "learning_rate": 2.119024527556127e-05, "loss": 0.0151, "step": 71070 }, { "grad_norm": 0.20134927332401276, "learning_rate": 2.117673284075415e-05, "loss": 0.0123, "step": 71080 }, { "grad_norm": 0.19327686727046967, "learning_rate": 2.1163223558016165e-05, "loss": 0.0104, "step": 71090 }, { "grad_norm": 0.0980159267783165, "learning_rate": 2.1149717428824722e-05, "loss": 0.0137, "step": 71100 }, { "grad_norm": 0.36603599786758423, "learning_rate": 2.1136214454656784e-05, "loss": 0.0169, "step": 71110 }, { "grad_norm": 0.10037868469953537, "learning_rate": 2.1122714636989056e-05, "loss": 0.0108, "step": 71120 }, { "grad_norm": 0.24929778277873993, "learning_rate": 2.1109217977297823e-05, "loss": 0.0198, "step": 71130 }, { "grad_norm": 0.10924273729324341, "learning_rate": 2.1095724477059077e-05, "loss": 0.0096, "step": 71140 }, { "grad_norm": 0.07916398346424103, "learning_rate": 2.1082234137748434e-05, "loss": 0.0109, "step": 71150 }, { "grad_norm": 0.16148164868354797, "learning_rate": 2.106874696084122e-05, "loss": 0.0108, "step": 71160 }, { "grad_norm": 0.33692726492881775, "learning_rate": 2.1055262947812315e-05, "loss": 0.0121, "step": 71170 }, { "grad_norm": 0.15041892230510712, "learning_rate": 2.104178210013636e-05, "loss": 0.0171, "step": 71180 }, { "grad_norm": 0.2463308572769165, "learning_rate": 2.1028304419287554e-05, "loss": 0.0118, "step": 71190 }, { "grad_norm": 0.19309690594673157, "learning_rate": 2.101482990673982e-05, "loss": 0.0128, "step": 71200 }, { "grad_norm": 0.22232691943645477, "learning_rate": 2.1001358563966707e-05, "loss": 0.0113, "step": 71210 }, { "grad_norm": 0.17210747301578522, "learning_rate": 2.098789039244144e-05, "loss": 0.0115, "step": 71220 }, { "grad_norm": 0.20024138689041138, "learning_rate": 2.097442539363685e-05, "loss": 0.0115, "step": 71230 }, { "grad_norm": 0.18048453330993652, "learning_rate": 2.0960963569025448e-05, "loss": 0.0151, "step": 71240 }, { "grad_norm": 0.34705501794815063, "learning_rate": 2.0947504920079435e-05, "loss": 0.0185, "step": 71250 }, { "grad_norm": 0.08490309864282608, "learning_rate": 2.093404944827059e-05, "loss": 0.0077, "step": 71260 }, { "grad_norm": 0.12950102984905243, "learning_rate": 2.0920597155070394e-05, "loss": 0.013, "step": 71270 }, { "grad_norm": 0.2675205171108246, "learning_rate": 2.090714804194997e-05, "loss": 0.0195, "step": 71280 }, { "grad_norm": 0.25124645233154297, "learning_rate": 2.0893702110380114e-05, "loss": 0.0165, "step": 71290 }, { "grad_norm": 0.1631290167570114, "learning_rate": 2.0880259361831216e-05, "loss": 0.0156, "step": 71300 }, { "grad_norm": 0.23073652386665344, "learning_rate": 2.0866819797773397e-05, "loss": 0.0112, "step": 71310 }, { "grad_norm": 0.2259979546070099, "learning_rate": 2.085338341967634e-05, "loss": 0.0118, "step": 71320 }, { "grad_norm": 0.1963471919298172, "learning_rate": 2.083995022900946e-05, "loss": 0.0198, "step": 71330 }, { "grad_norm": 0.37047940492630005, "learning_rate": 2.0826520227241768e-05, "loss": 0.0147, "step": 71340 }, { "grad_norm": 0.14432169497013092, "learning_rate": 2.081309341584199e-05, "loss": 0.0167, "step": 71350 }, { "grad_norm": 0.24435390532016754, "learning_rate": 2.0799669796278413e-05, "loss": 0.0137, "step": 71360 }, { "grad_norm": 0.329897940158844, "learning_rate": 2.078624937001906e-05, "loss": 0.0158, "step": 71370 }, { "grad_norm": 0.18067000806331635, "learning_rate": 2.0772832138531544e-05, "loss": 0.0303, "step": 71380 }, { "grad_norm": 0.3509736657142639, "learning_rate": 2.0759418103283158e-05, "loss": 0.0149, "step": 71390 }, { "grad_norm": 0.15161381661891937, "learning_rate": 2.0746007265740846e-05, "loss": 0.0154, "step": 71400 }, { "grad_norm": 0.696185827255249, "learning_rate": 2.0732599627371215e-05, "loss": 0.013, "step": 71410 }, { "grad_norm": 0.1558901071548462, "learning_rate": 2.0719195189640468e-05, "loss": 0.0131, "step": 71420 }, { "grad_norm": 0.08336204290390015, "learning_rate": 2.0705793954014528e-05, "loss": 0.0075, "step": 71430 }, { "grad_norm": 0.3139829635620117, "learning_rate": 2.06923959219589e-05, "loss": 0.0127, "step": 71440 }, { "grad_norm": 0.4133165180683136, "learning_rate": 2.0679001094938793e-05, "loss": 0.0143, "step": 71450 }, { "grad_norm": 0.20802710950374603, "learning_rate": 2.066560947441904e-05, "loss": 0.0082, "step": 71460 }, { "grad_norm": 0.0875164270401001, "learning_rate": 2.0652221061864152e-05, "loss": 0.0146, "step": 71470 }, { "grad_norm": 0.2636067271232605, "learning_rate": 2.0638835858738226e-05, "loss": 0.015, "step": 71480 }, { "grad_norm": 0.3025297522544861, "learning_rate": 2.0625453866505068e-05, "loss": 0.0134, "step": 71490 }, { "grad_norm": 0.35917750000953674, "learning_rate": 2.0612075086628123e-05, "loss": 0.019, "step": 71500 }, { "grad_norm": 0.35174131393432617, "learning_rate": 2.059869952057045e-05, "loss": 0.021, "step": 71510 }, { "grad_norm": 0.34017521142959595, "learning_rate": 2.0585327169794793e-05, "loss": 0.0142, "step": 71520 }, { "grad_norm": 0.5335085988044739, "learning_rate": 2.0571958035763534e-05, "loss": 0.018, "step": 71530 }, { "grad_norm": 0.27619117498397827, "learning_rate": 2.0558592119938715e-05, "loss": 0.0089, "step": 71540 }, { "grad_norm": 0.11849082261323929, "learning_rate": 2.054522942378198e-05, "loss": 0.0101, "step": 71550 }, { "grad_norm": 0.5961951613426208, "learning_rate": 2.053186994875469e-05, "loss": 0.0076, "step": 71560 }, { "grad_norm": 0.12071786820888519, "learning_rate": 2.0518513696317775e-05, "loss": 0.0086, "step": 71570 }, { "grad_norm": 0.24511149525642395, "learning_rate": 2.050516066793188e-05, "loss": 0.0245, "step": 71580 }, { "grad_norm": 0.2194056212902069, "learning_rate": 2.0491810865057276e-05, "loss": 0.0154, "step": 71590 }, { "grad_norm": 0.21989287436008453, "learning_rate": 2.0478464289153882e-05, "loss": 0.0105, "step": 71600 }, { "grad_norm": 0.08947382122278214, "learning_rate": 2.0465120941681232e-05, "loss": 0.0108, "step": 71610 }, { "grad_norm": 0.15981589257717133, "learning_rate": 2.0451780824098566e-05, "loss": 0.0143, "step": 71620 }, { "grad_norm": 0.07931744307279587, "learning_rate": 2.043844393786471e-05, "loss": 0.0098, "step": 71630 }, { "grad_norm": 0.06729115545749664, "learning_rate": 2.0425110284438193e-05, "loss": 0.0112, "step": 71640 }, { "grad_norm": 0.15728168189525604, "learning_rate": 2.041177986527712e-05, "loss": 0.0118, "step": 71650 }, { "grad_norm": 0.17486290633678436, "learning_rate": 2.0398452681839347e-05, "loss": 0.0111, "step": 71660 }, { "grad_norm": 0.6243111491203308, "learning_rate": 2.0385128735582272e-05, "loss": 0.0329, "step": 71670 }, { "grad_norm": 0.4505367577075958, "learning_rate": 2.0371808027963003e-05, "loss": 0.0178, "step": 71680 }, { "grad_norm": 0.3576127588748932, "learning_rate": 2.035849056043825e-05, "loss": 0.0129, "step": 71690 }, { "grad_norm": 0.06501541286706924, "learning_rate": 2.03451763344644e-05, "loss": 0.008, "step": 71700 }, { "grad_norm": 0.1506902426481247, "learning_rate": 2.0331865351497482e-05, "loss": 0.0126, "step": 71710 }, { "grad_norm": 0.3002317547798157, "learning_rate": 2.0318557612993185e-05, "loss": 0.0148, "step": 71720 }, { "grad_norm": 0.2842172682285309, "learning_rate": 2.0305253120406782e-05, "loss": 0.0144, "step": 71730 }, { "grad_norm": 0.32239624857902527, "learning_rate": 2.029195187519325e-05, "loss": 0.0104, "step": 71740 }, { "grad_norm": 0.12401261925697327, "learning_rate": 2.0278653878807217e-05, "loss": 0.0162, "step": 71750 }, { "grad_norm": 0.33714425563812256, "learning_rate": 2.0265359132702892e-05, "loss": 0.022, "step": 71760 }, { "grad_norm": 0.4563133120536804, "learning_rate": 2.0252067638334182e-05, "loss": 0.0172, "step": 71770 }, { "grad_norm": 0.0941573828458786, "learning_rate": 2.023877939715464e-05, "loss": 0.0103, "step": 71780 }, { "grad_norm": 0.11200758814811707, "learning_rate": 2.0225494410617448e-05, "loss": 0.0162, "step": 71790 }, { "grad_norm": 0.18300795555114746, "learning_rate": 2.02122126801754e-05, "loss": 0.0157, "step": 71800 }, { "grad_norm": 0.12264259159564972, "learning_rate": 2.0198934207281018e-05, "loss": 0.0162, "step": 71810 }, { "grad_norm": 0.16946600377559662, "learning_rate": 2.0185658993386357e-05, "loss": 0.0127, "step": 71820 }, { "grad_norm": 0.1741315722465515, "learning_rate": 2.0172387039943226e-05, "loss": 0.0093, "step": 71830 }, { "grad_norm": 0.1804651916027069, "learning_rate": 2.0159118348402968e-05, "loss": 0.0216, "step": 71840 }, { "grad_norm": 0.08428497612476349, "learning_rate": 2.0145852920216697e-05, "loss": 0.0162, "step": 71850 }, { "grad_norm": 0.0632036030292511, "learning_rate": 2.0132590756835045e-05, "loss": 0.0148, "step": 71860 }, { "grad_norm": 0.2836243212223053, "learning_rate": 2.0119331859708386e-05, "loss": 0.0244, "step": 71870 }, { "grad_norm": 0.20558638870716095, "learning_rate": 2.0106076230286654e-05, "loss": 0.0178, "step": 71880 }, { "grad_norm": 0.21487461030483246, "learning_rate": 2.0092823870019488e-05, "loss": 0.0122, "step": 71890 }, { "grad_norm": 0.15822118520736694, "learning_rate": 2.0079574780356115e-05, "loss": 0.0147, "step": 71900 }, { "grad_norm": 0.08577775955200195, "learning_rate": 2.0066328962745486e-05, "loss": 0.0101, "step": 71910 }, { "grad_norm": 0.08028092235326767, "learning_rate": 2.00530864186361e-05, "loss": 0.0096, "step": 71920 }, { "grad_norm": 0.14357325434684753, "learning_rate": 2.003984714947618e-05, "loss": 0.0158, "step": 71930 }, { "grad_norm": 0.3487600088119507, "learning_rate": 2.0026611156713515e-05, "loss": 0.0156, "step": 71940 }, { "grad_norm": 0.6213803887367249, "learning_rate": 2.0013378441795584e-05, "loss": 0.0192, "step": 71950 }, { "grad_norm": 0.2792021632194519, "learning_rate": 2.0000149006169506e-05, "loss": 0.0085, "step": 71960 }, { "grad_norm": 0.2180452048778534, "learning_rate": 1.9986922851282048e-05, "loss": 0.0198, "step": 71970 }, { "grad_norm": 0.14085052907466888, "learning_rate": 1.997369997857956e-05, "loss": 0.0142, "step": 71980 }, { "grad_norm": 0.19211970269680023, "learning_rate": 1.9960480389508103e-05, "loss": 0.0103, "step": 71990 }, { "grad_norm": 0.2425963282585144, "learning_rate": 1.9947264085513363e-05, "loss": 0.0202, "step": 72000 }, { "grad_norm": 0.2508067190647125, "learning_rate": 1.993405106804062e-05, "loss": 0.0204, "step": 72010 }, { "grad_norm": 0.14218658208847046, "learning_rate": 1.9920841338534867e-05, "loss": 0.0182, "step": 72020 }, { "grad_norm": 0.3356335163116455, "learning_rate": 1.9907634898440646e-05, "loss": 0.0131, "step": 72030 }, { "grad_norm": 0.12208999693393707, "learning_rate": 1.989443174920227e-05, "loss": 0.0157, "step": 72040 }, { "grad_norm": 0.31841176748275757, "learning_rate": 1.988123189226355e-05, "loss": 0.0232, "step": 72050 }, { "grad_norm": 0.1102108359336853, "learning_rate": 1.986803532906805e-05, "loss": 0.0121, "step": 72060 }, { "grad_norm": 0.10481062531471252, "learning_rate": 1.985484206105889e-05, "loss": 0.0094, "step": 72070 }, { "grad_norm": 0.06519437581300735, "learning_rate": 1.9841652089678887e-05, "loss": 0.0071, "step": 72080 }, { "grad_norm": 0.238492950797081, "learning_rate": 1.9828465416370435e-05, "loss": 0.0157, "step": 72090 }, { "grad_norm": 0.13950715959072113, "learning_rate": 1.9815282042575685e-05, "loss": 0.0139, "step": 72100 }, { "grad_norm": 0.10346372425556183, "learning_rate": 1.9802101969736285e-05, "loss": 0.0202, "step": 72110 }, { "grad_norm": 0.2235233187675476, "learning_rate": 1.9788925199293633e-05, "loss": 0.0146, "step": 72120 }, { "grad_norm": 0.14814770221710205, "learning_rate": 1.977575173268868e-05, "loss": 0.0203, "step": 72130 }, { "grad_norm": 0.39479348063468933, "learning_rate": 1.9762581571362094e-05, "loss": 0.0106, "step": 72140 }, { "grad_norm": 0.3940846920013428, "learning_rate": 1.9749414716754112e-05, "loss": 0.0156, "step": 72150 }, { "grad_norm": 0.20997431874275208, "learning_rate": 1.973625117030465e-05, "loss": 0.0144, "step": 72160 }, { "grad_norm": 0.5222761034965515, "learning_rate": 1.972309093345326e-05, "loss": 0.0191, "step": 72170 }, { "grad_norm": 0.323606014251709, "learning_rate": 1.970993400763915e-05, "loss": 0.0115, "step": 72180 }, { "grad_norm": 0.1586272269487381, "learning_rate": 1.969678039430109e-05, "loss": 0.0123, "step": 72190 }, { "grad_norm": 0.43073761463165283, "learning_rate": 1.9683630094877564e-05, "loss": 0.0142, "step": 72200 }, { "grad_norm": 0.16045010089874268, "learning_rate": 1.9670483110806696e-05, "loss": 0.0079, "step": 72210 }, { "grad_norm": 0.15780788660049438, "learning_rate": 1.9657339443526172e-05, "loss": 0.0208, "step": 72220 }, { "grad_norm": 0.08784916996955872, "learning_rate": 1.964419909447339e-05, "loss": 0.0107, "step": 72230 }, { "grad_norm": 0.04979369416832924, "learning_rate": 1.963106206508536e-05, "loss": 0.0146, "step": 72240 }, { "grad_norm": 0.15465250611305237, "learning_rate": 1.9617928356798738e-05, "loss": 0.0095, "step": 72250 }, { "grad_norm": 0.06248732656240463, "learning_rate": 1.960479797104977e-05, "loss": 0.0146, "step": 72260 }, { "grad_norm": 0.1615913063287735, "learning_rate": 1.9591670909274418e-05, "loss": 0.0092, "step": 72270 }, { "grad_norm": 0.06763078272342682, "learning_rate": 1.9578547172908184e-05, "loss": 0.0148, "step": 72280 }, { "grad_norm": 0.07184702157974243, "learning_rate": 1.9565426763386325e-05, "loss": 0.0109, "step": 72290 }, { "grad_norm": 0.07478529959917068, "learning_rate": 1.9552309682143622e-05, "loss": 0.0125, "step": 72300 }, { "grad_norm": 0.2188858538866043, "learning_rate": 1.9539195930614574e-05, "loss": 0.0239, "step": 72310 }, { "grad_norm": 0.2923571765422821, "learning_rate": 1.952608551023324e-05, "loss": 0.021, "step": 72320 }, { "grad_norm": 0.0896189957857132, "learning_rate": 1.9512978422433397e-05, "loss": 0.015, "step": 72330 }, { "grad_norm": 0.08244165778160095, "learning_rate": 1.9499874668648376e-05, "loss": 0.01, "step": 72340 }, { "grad_norm": 0.15467621386051178, "learning_rate": 1.9486774250311213e-05, "loss": 0.0116, "step": 72350 }, { "grad_norm": 0.29816171526908875, "learning_rate": 1.9473677168854537e-05, "loss": 0.0091, "step": 72360 }, { "grad_norm": 0.10311846435070038, "learning_rate": 1.946058342571065e-05, "loss": 0.0128, "step": 72370 }, { "grad_norm": 0.1308605968952179, "learning_rate": 1.9447493022311424e-05, "loss": 0.0105, "step": 72380 }, { "grad_norm": 0.22702310979366302, "learning_rate": 1.9434405960088442e-05, "loss": 0.0106, "step": 72390 }, { "grad_norm": 0.22610178589820862, "learning_rate": 1.9421322240472857e-05, "loss": 0.018, "step": 72400 }, { "grad_norm": 0.18177704513072968, "learning_rate": 1.9408241864895494e-05, "loss": 0.0123, "step": 72410 }, { "grad_norm": 0.32301411032676697, "learning_rate": 1.939516483478681e-05, "loss": 0.0174, "step": 72420 }, { "grad_norm": 0.11872677505016327, "learning_rate": 1.9382091151576902e-05, "loss": 0.0158, "step": 72430 }, { "grad_norm": 0.24179257452487946, "learning_rate": 1.9369020816695454e-05, "loss": 0.0184, "step": 72440 }, { "grad_norm": 0.2051495611667633, "learning_rate": 1.935595383157184e-05, "loss": 0.0124, "step": 72450 }, { "grad_norm": 0.15783676505088806, "learning_rate": 1.9342890197635057e-05, "loss": 0.0083, "step": 72460 }, { "grad_norm": 0.2938153147697449, "learning_rate": 1.9329829916313685e-05, "loss": 0.0188, "step": 72470 }, { "grad_norm": 0.0932626873254776, "learning_rate": 1.9316772989036013e-05, "loss": 0.0129, "step": 72480 }, { "grad_norm": 0.28863224387168884, "learning_rate": 1.930371941722991e-05, "loss": 0.0127, "step": 72490 }, { "grad_norm": 0.25302714109420776, "learning_rate": 1.929066920232292e-05, "loss": 0.0097, "step": 72500 }, { "grad_norm": 0.2888461649417877, "learning_rate": 1.9277622345742152e-05, "loss": 0.0166, "step": 72510 }, { "grad_norm": 0.1454564929008484, "learning_rate": 1.926457884891444e-05, "loss": 0.0126, "step": 72520 }, { "grad_norm": 0.2310151606798172, "learning_rate": 1.9251538713266147e-05, "loss": 0.0136, "step": 72530 }, { "grad_norm": 0.39623555541038513, "learning_rate": 1.9238501940223358e-05, "loss": 0.0114, "step": 72540 }, { "grad_norm": 0.1470532864332199, "learning_rate": 1.922546853121175e-05, "loss": 0.0117, "step": 72550 }, { "grad_norm": 0.12438087165355682, "learning_rate": 1.9212438487656647e-05, "loss": 0.0249, "step": 72560 }, { "grad_norm": 0.10317351669073105, "learning_rate": 1.9199411810982965e-05, "loss": 0.0187, "step": 72570 }, { "grad_norm": 0.17806287109851837, "learning_rate": 1.918638850261532e-05, "loss": 0.0145, "step": 72580 }, { "grad_norm": 0.1112140491604805, "learning_rate": 1.9173368563977885e-05, "loss": 0.0081, "step": 72590 }, { "grad_norm": 0.43911105394363403, "learning_rate": 1.9160351996494513e-05, "loss": 0.0142, "step": 72600 }, { "grad_norm": 0.10474516451358795, "learning_rate": 1.9147338801588677e-05, "loss": 0.0162, "step": 72610 }, { "grad_norm": 0.22850768268108368, "learning_rate": 1.913432898068351e-05, "loss": 0.0167, "step": 72620 }, { "grad_norm": 0.19992297887802124, "learning_rate": 1.912132253520169e-05, "loss": 0.0147, "step": 72630 }, { "grad_norm": 0.18232707679271698, "learning_rate": 1.910831946656564e-05, "loss": 0.0149, "step": 72640 }, { "grad_norm": 0.07103149592876434, "learning_rate": 1.909531977619731e-05, "loss": 0.0091, "step": 72650 }, { "grad_norm": 0.32298457622528076, "learning_rate": 1.908232346551834e-05, "loss": 0.0106, "step": 72660 }, { "grad_norm": 0.31638363003730774, "learning_rate": 1.9069330535949993e-05, "loss": 0.0148, "step": 72670 }, { "grad_norm": 0.21417087316513062, "learning_rate": 1.9056340988913167e-05, "loss": 0.0128, "step": 72680 }, { "grad_norm": 0.1105995699763298, "learning_rate": 1.9043354825828357e-05, "loss": 0.0132, "step": 72690 }, { "grad_norm": 0.08852893859148026, "learning_rate": 1.9030372048115724e-05, "loss": 0.0118, "step": 72700 }, { "grad_norm": 0.14728733897209167, "learning_rate": 1.9017392657195056e-05, "loss": 0.0248, "step": 72710 }, { "grad_norm": 0.2152918428182602, "learning_rate": 1.9004416654485724e-05, "loss": 0.0164, "step": 72720 }, { "grad_norm": 0.2509857714176178, "learning_rate": 1.8991444041406785e-05, "loss": 0.0182, "step": 72730 }, { "grad_norm": 0.12397052347660065, "learning_rate": 1.8978474819376912e-05, "loss": 0.0158, "step": 72740 }, { "grad_norm": 0.08565467596054077, "learning_rate": 1.8965508989814406e-05, "loss": 0.0127, "step": 72750 }, { "grad_norm": 0.1755804717540741, "learning_rate": 1.8952546554137158e-05, "loss": 0.0101, "step": 72760 }, { "grad_norm": 0.22795911133289337, "learning_rate": 1.8939587513762757e-05, "loss": 0.0134, "step": 72770 }, { "grad_norm": 0.22414612770080566, "learning_rate": 1.892663187010835e-05, "loss": 0.0129, "step": 72780 }, { "grad_norm": 0.18473728001117706, "learning_rate": 1.891367962459077e-05, "loss": 0.0193, "step": 72790 }, { "grad_norm": 0.14072227478027344, "learning_rate": 1.890073077862644e-05, "loss": 0.008, "step": 72800 }, { "grad_norm": 0.14680986106395721, "learning_rate": 1.888778533363146e-05, "loss": 0.0092, "step": 72810 }, { "grad_norm": 0.045560095459222794, "learning_rate": 1.8874843291021478e-05, "loss": 0.0092, "step": 72820 }, { "grad_norm": 0.15958300232887268, "learning_rate": 1.8861904652211853e-05, "loss": 0.0156, "step": 72830 }, { "grad_norm": 0.2115592509508133, "learning_rate": 1.8848969418617507e-05, "loss": 0.0105, "step": 72840 }, { "grad_norm": 0.3120706379413605, "learning_rate": 1.8836037591653044e-05, "loss": 0.0099, "step": 72850 }, { "grad_norm": 0.11922120302915573, "learning_rate": 1.882310917273262e-05, "loss": 0.0092, "step": 72860 }, { "grad_norm": 0.27991345524787903, "learning_rate": 1.881018416327014e-05, "loss": 0.0105, "step": 72870 }, { "grad_norm": 0.12861980497837067, "learning_rate": 1.8797262564679014e-05, "loss": 0.0162, "step": 72880 }, { "grad_norm": 0.3010316789150238, "learning_rate": 1.8784344378372355e-05, "loss": 0.0109, "step": 72890 }, { "grad_norm": 0.1595509946346283, "learning_rate": 1.8771429605762843e-05, "loss": 0.0089, "step": 72900 }, { "grad_norm": 0.12273737788200378, "learning_rate": 1.8758518248262835e-05, "loss": 0.0081, "step": 72910 }, { "grad_norm": 0.2475913017988205, "learning_rate": 1.8745610307284305e-05, "loss": 0.0181, "step": 72920 }, { "grad_norm": 0.13276740908622742, "learning_rate": 1.8732705784238862e-05, "loss": 0.0183, "step": 72930 }, { "grad_norm": 0.11844808608293533, "learning_rate": 1.8719804680537682e-05, "loss": 0.0134, "step": 72940 }, { "grad_norm": 0.1493924856185913, "learning_rate": 1.870690699759164e-05, "loss": 0.0099, "step": 72950 }, { "grad_norm": 0.1284559965133667, "learning_rate": 1.8694012736811213e-05, "loss": 0.0151, "step": 72960 }, { "grad_norm": 0.33620545268058777, "learning_rate": 1.8681121899606463e-05, "loss": 0.0117, "step": 72970 }, { "grad_norm": 0.1030551865696907, "learning_rate": 1.8668234487387144e-05, "loss": 0.0091, "step": 72980 }, { "grad_norm": 0.16553129255771637, "learning_rate": 1.8655350501562592e-05, "loss": 0.0166, "step": 72990 }, { "grad_norm": 0.16442103683948517, "learning_rate": 1.8642469943541796e-05, "loss": 0.0216, "step": 73000 }, { "grad_norm": 0.12651877105236053, "learning_rate": 1.8629592814733328e-05, "loss": 0.0081, "step": 73010 }, { "grad_norm": 0.1600286215543747, "learning_rate": 1.8616719116545438e-05, "loss": 0.0223, "step": 73020 }, { "grad_norm": 0.08771967142820358, "learning_rate": 1.8603848850385952e-05, "loss": 0.0097, "step": 73030 }, { "grad_norm": 0.1619168370962143, "learning_rate": 1.859098201766236e-05, "loss": 0.0177, "step": 73040 }, { "grad_norm": 0.21374067664146423, "learning_rate": 1.8578118619781726e-05, "loss": 0.0175, "step": 73050 }, { "grad_norm": 0.10525400936603546, "learning_rate": 1.856525865815083e-05, "loss": 0.0113, "step": 73060 }, { "grad_norm": 0.2384423166513443, "learning_rate": 1.8552402134175966e-05, "loss": 0.0088, "step": 73070 }, { "grad_norm": 0.1056026965379715, "learning_rate": 1.8539549049263143e-05, "loss": 0.0089, "step": 73080 }, { "grad_norm": 0.07531873136758804, "learning_rate": 1.8526699404817917e-05, "loss": 0.0139, "step": 73090 }, { "grad_norm": 0.1510886251926422, "learning_rate": 1.8513853202245546e-05, "loss": 0.0169, "step": 73100 }, { "grad_norm": 0.15557843446731567, "learning_rate": 1.850101044295081e-05, "loss": 0.0099, "step": 73110 }, { "grad_norm": 0.08956768363714218, "learning_rate": 1.848817112833825e-05, "loss": 0.0096, "step": 73120 }, { "grad_norm": 0.12936295568943024, "learning_rate": 1.8475335259811905e-05, "loss": 0.0129, "step": 73130 }, { "grad_norm": 0.22584319114685059, "learning_rate": 1.8462502838775508e-05, "loss": 0.0238, "step": 73140 }, { "grad_norm": 0.15250100195407867, "learning_rate": 1.8449673866632378e-05, "loss": 0.0147, "step": 73150 }, { "grad_norm": 0.1267584264278412, "learning_rate": 1.8436848344785473e-05, "loss": 0.0136, "step": 73160 }, { "grad_norm": 0.23604263365268707, "learning_rate": 1.8424026274637378e-05, "loss": 0.0173, "step": 73170 }, { "grad_norm": 0.15514406561851501, "learning_rate": 1.8411207657590312e-05, "loss": 0.0174, "step": 73180 }, { "grad_norm": 0.1777709722518921, "learning_rate": 1.839839249504607e-05, "loss": 0.016, "step": 73190 }, { "grad_norm": 0.4102703332901001, "learning_rate": 1.8385580788406108e-05, "loss": 0.0184, "step": 73200 }, { "grad_norm": 0.11385077238082886, "learning_rate": 1.837277253907152e-05, "loss": 0.0097, "step": 73210 }, { "grad_norm": 0.07303635776042938, "learning_rate": 1.8359967748442957e-05, "loss": 0.0152, "step": 73220 }, { "grad_norm": 0.23611164093017578, "learning_rate": 1.8347166417920774e-05, "loss": 0.0098, "step": 73230 }, { "grad_norm": 0.32077300548553467, "learning_rate": 1.833436854890484e-05, "loss": 0.0161, "step": 73240 }, { "grad_norm": 0.14448092877864838, "learning_rate": 1.8321574142794796e-05, "loss": 0.0147, "step": 73250 }, { "grad_norm": 0.243039071559906, "learning_rate": 1.8308783200989748e-05, "loss": 0.0109, "step": 73260 }, { "grad_norm": 0.16854415833950043, "learning_rate": 1.829599572488855e-05, "loss": 0.014, "step": 73270 }, { "grad_norm": 0.2156752347946167, "learning_rate": 1.828321171588957e-05, "loss": 0.0105, "step": 73280 }, { "grad_norm": 0.14503775537014008, "learning_rate": 1.827043117539089e-05, "loss": 0.0108, "step": 73290 }, { "grad_norm": 0.3462335765361786, "learning_rate": 1.8257654104790116e-05, "loss": 0.0111, "step": 73300 }, { "grad_norm": 0.07981288433074951, "learning_rate": 1.82448805054846e-05, "loss": 0.0128, "step": 73310 }, { "grad_norm": 0.04571323096752167, "learning_rate": 1.823211037887119e-05, "loss": 0.0186, "step": 73320 }, { "grad_norm": 0.2679240107536316, "learning_rate": 1.821934372634645e-05, "loss": 0.0144, "step": 73330 }, { "grad_norm": 0.2226567566394806, "learning_rate": 1.8206580549306472e-05, "loss": 0.0093, "step": 73340 }, { "grad_norm": 0.08764299005270004, "learning_rate": 1.819382084914707e-05, "loss": 0.0167, "step": 73350 }, { "grad_norm": 0.15431112051010132, "learning_rate": 1.8181064627263577e-05, "loss": 0.0071, "step": 73360 }, { "grad_norm": 0.18898329138755798, "learning_rate": 1.8168311885051022e-05, "loss": 0.0104, "step": 73370 }, { "grad_norm": 0.16437548398971558, "learning_rate": 1.815556262390402e-05, "loss": 0.0145, "step": 73380 }, { "grad_norm": 0.6268168091773987, "learning_rate": 1.8142816845216826e-05, "loss": 0.0127, "step": 73390 }, { "grad_norm": 0.17171066999435425, "learning_rate": 1.8130074550383276e-05, "loss": 0.0124, "step": 73400 }, { "grad_norm": 0.14160914719104767, "learning_rate": 1.8117335740796848e-05, "loss": 0.0128, "step": 73410 }, { "grad_norm": 0.1198921650648117, "learning_rate": 1.810460041785067e-05, "loss": 0.0109, "step": 73420 }, { "grad_norm": 0.19857154786586761, "learning_rate": 1.809186858293742e-05, "loss": 0.0221, "step": 73430 }, { "grad_norm": 0.23320384323596954, "learning_rate": 1.8079140237449448e-05, "loss": 0.0129, "step": 73440 }, { "grad_norm": 0.43257686495780945, "learning_rate": 1.8066415382778707e-05, "loss": 0.0182, "step": 73450 }, { "grad_norm": 0.1296750158071518, "learning_rate": 1.8053694020316786e-05, "loss": 0.0138, "step": 73460 }, { "grad_norm": 0.12866820394992828, "learning_rate": 1.8040976151454837e-05, "loss": 0.0126, "step": 73470 }, { "grad_norm": 0.13851378858089447, "learning_rate": 1.8028261777583704e-05, "loss": 0.0196, "step": 73480 }, { "grad_norm": 0.06411565840244293, "learning_rate": 1.8015550900093758e-05, "loss": 0.0106, "step": 73490 }, { "grad_norm": 0.2532576024532318, "learning_rate": 1.800284352037512e-05, "loss": 0.0132, "step": 73500 }, { "grad_norm": 0.19950078427791595, "learning_rate": 1.799013963981738e-05, "loss": 0.0144, "step": 73510 }, { "grad_norm": 0.18204998970031738, "learning_rate": 1.797743925980987e-05, "loss": 0.0201, "step": 73520 }, { "grad_norm": 0.09934794157743454, "learning_rate": 1.7964742381741435e-05, "loss": 0.0095, "step": 73530 }, { "grad_norm": 0.11837964504957199, "learning_rate": 1.7952049007000627e-05, "loss": 0.0177, "step": 73540 }, { "grad_norm": 0.2915451228618622, "learning_rate": 1.7939359136975535e-05, "loss": 0.0196, "step": 73550 }, { "grad_norm": 0.2440340369939804, "learning_rate": 1.7926672773053932e-05, "loss": 0.0142, "step": 73560 }, { "grad_norm": 0.20791210234165192, "learning_rate": 1.791398991662317e-05, "loss": 0.0212, "step": 73570 }, { "grad_norm": 0.19113938510417938, "learning_rate": 1.7901310569070244e-05, "loss": 0.0203, "step": 73580 }, { "grad_norm": 0.07517114281654358, "learning_rate": 1.7888634731781724e-05, "loss": 0.01, "step": 73590 }, { "grad_norm": 0.14987672865390778, "learning_rate": 1.7875962406143844e-05, "loss": 0.0103, "step": 73600 }, { "grad_norm": 0.06914553791284561, "learning_rate": 1.7863293593542403e-05, "loss": 0.0241, "step": 73610 }, { "grad_norm": 0.0657847449183464, "learning_rate": 1.7850628295362853e-05, "loss": 0.0158, "step": 73620 }, { "grad_norm": 0.1907169669866562, "learning_rate": 1.7837966512990263e-05, "loss": 0.0132, "step": 73630 }, { "grad_norm": 0.10986486077308655, "learning_rate": 1.782530824780931e-05, "loss": 0.0129, "step": 73640 }, { "grad_norm": 0.25167956948280334, "learning_rate": 1.7812653501204263e-05, "loss": 0.0157, "step": 73650 }, { "grad_norm": 0.09754297137260437, "learning_rate": 1.780000227455903e-05, "loss": 0.0153, "step": 73660 }, { "grad_norm": 0.10383031517267227, "learning_rate": 1.778735456925716e-05, "loss": 0.013, "step": 73670 }, { "grad_norm": 0.16713981330394745, "learning_rate": 1.7774710386681737e-05, "loss": 0.0119, "step": 73680 }, { "grad_norm": 0.11391230672597885, "learning_rate": 1.7762069728215545e-05, "loss": 0.0122, "step": 73690 }, { "grad_norm": 0.35515838861465454, "learning_rate": 1.7749432595240932e-05, "loss": 0.0184, "step": 73700 }, { "grad_norm": 0.23799511790275574, "learning_rate": 1.773679898913991e-05, "loss": 0.0215, "step": 73710 }, { "grad_norm": 0.23554329574108124, "learning_rate": 1.7724168911294015e-05, "loss": 0.0109, "step": 73720 }, { "grad_norm": 0.06838792562484741, "learning_rate": 1.7711542363084504e-05, "loss": 0.0104, "step": 73730 }, { "grad_norm": 0.29345670342445374, "learning_rate": 1.7698919345892158e-05, "loss": 0.0137, "step": 73740 }, { "grad_norm": 0.2653143107891083, "learning_rate": 1.768629986109743e-05, "loss": 0.0172, "step": 73750 }, { "grad_norm": 0.2637004554271698, "learning_rate": 1.7673683910080357e-05, "loss": 0.012, "step": 73760 }, { "grad_norm": 0.3656712472438812, "learning_rate": 1.7661071494220638e-05, "loss": 0.0156, "step": 73770 }, { "grad_norm": 0.5401975512504578, "learning_rate": 1.7648462614897498e-05, "loss": 0.0136, "step": 73780 }, { "grad_norm": 0.18688304722309113, "learning_rate": 1.7635857273489863e-05, "loss": 0.0138, "step": 73790 }, { "grad_norm": 0.3681979477405548, "learning_rate": 1.7623255471376198e-05, "loss": 0.0171, "step": 73800 }, { "grad_norm": 0.35671916604042053, "learning_rate": 1.7610657209934646e-05, "loss": 0.0231, "step": 73810 }, { "grad_norm": 0.14968326687812805, "learning_rate": 1.7598062490542915e-05, "loss": 0.0109, "step": 73820 }, { "grad_norm": 0.3443372845649719, "learning_rate": 1.7585471314578377e-05, "loss": 0.018, "step": 73830 }, { "grad_norm": 0.19203346967697144, "learning_rate": 1.7572883683417945e-05, "loss": 0.0157, "step": 73840 }, { "grad_norm": 0.1866552233695984, "learning_rate": 1.7560299598438212e-05, "loss": 0.0175, "step": 73850 }, { "grad_norm": 0.15631724894046783, "learning_rate": 1.7547719061015334e-05, "loss": 0.0118, "step": 73860 }, { "grad_norm": 0.09948249906301498, "learning_rate": 1.75351420725251e-05, "loss": 0.0154, "step": 73870 }, { "grad_norm": 0.15094713866710663, "learning_rate": 1.7522568634342923e-05, "loss": 0.0134, "step": 73880 }, { "grad_norm": 0.051664840430021286, "learning_rate": 1.7509998747843825e-05, "loss": 0.0062, "step": 73890 }, { "grad_norm": 0.13897062838077545, "learning_rate": 1.7497432414402403e-05, "loss": 0.0097, "step": 73900 }, { "grad_norm": 0.1848476082086563, "learning_rate": 1.7484869635392903e-05, "loss": 0.0098, "step": 73910 }, { "grad_norm": 0.17579394578933716, "learning_rate": 1.747231041218919e-05, "loss": 0.0185, "step": 73920 }, { "grad_norm": 0.5090737342834473, "learning_rate": 1.7459754746164692e-05, "loss": 0.024, "step": 73930 }, { "grad_norm": 0.1587247997522354, "learning_rate": 1.744720263869249e-05, "loss": 0.0165, "step": 73940 }, { "grad_norm": 0.08006113767623901, "learning_rate": 1.7434654091145263e-05, "loss": 0.0114, "step": 73950 }, { "grad_norm": 0.274905800819397, "learning_rate": 1.742210910489532e-05, "loss": 0.0136, "step": 73960 }, { "grad_norm": 0.18944838643074036, "learning_rate": 1.7409567681314532e-05, "loss": 0.0121, "step": 73970 }, { "grad_norm": 0.3306540250778198, "learning_rate": 1.7397029821774434e-05, "loss": 0.0132, "step": 73980 }, { "grad_norm": 0.09400864690542221, "learning_rate": 1.7384495527646126e-05, "loss": 0.0128, "step": 73990 }, { "grad_norm": 0.1832100749015808, "learning_rate": 1.737196480030035e-05, "loss": 0.0083, "step": 74000 }, { "grad_norm": 0.07027330994606018, "learning_rate": 1.7359437641107447e-05, "loss": 0.0154, "step": 74010 }, { "grad_norm": 0.22889545559883118, "learning_rate": 1.7346914051437387e-05, "loss": 0.0207, "step": 74020 }, { "grad_norm": 0.23365093767642975, "learning_rate": 1.733439403265969e-05, "loss": 0.0125, "step": 74030 }, { "grad_norm": 0.23221221566200256, "learning_rate": 1.732187758614357e-05, "loss": 0.0207, "step": 74040 }, { "grad_norm": 0.14673401415348053, "learning_rate": 1.730936471325777e-05, "loss": 0.0126, "step": 74050 }, { "grad_norm": 0.030329974368214607, "learning_rate": 1.7296855415370718e-05, "loss": 0.0177, "step": 74060 }, { "grad_norm": 0.1651882380247116, "learning_rate": 1.728434969385035e-05, "loss": 0.0115, "step": 74070 }, { "grad_norm": 0.15264299511909485, "learning_rate": 1.7271847550064346e-05, "loss": 0.0103, "step": 74080 }, { "grad_norm": 0.16805489361286163, "learning_rate": 1.725934898537987e-05, "loss": 0.018, "step": 74090 }, { "grad_norm": 0.18870903551578522, "learning_rate": 1.7246854001163783e-05, "loss": 0.0099, "step": 74100 }, { "grad_norm": 0.05144333839416504, "learning_rate": 1.7234362598782476e-05, "loss": 0.0097, "step": 74110 }, { "grad_norm": 0.05090111866593361, "learning_rate": 1.722187477960202e-05, "loss": 0.0137, "step": 74120 }, { "grad_norm": 0.12588372826576233, "learning_rate": 1.7209390544988057e-05, "loss": 0.0197, "step": 74130 }, { "grad_norm": 0.3042561113834381, "learning_rate": 1.7196909896305856e-05, "loss": 0.0234, "step": 74140 }, { "grad_norm": 0.2113741934299469, "learning_rate": 1.718443283492026e-05, "loss": 0.02, "step": 74150 }, { "grad_norm": 0.4116208255290985, "learning_rate": 1.7171959362195746e-05, "loss": 0.0187, "step": 74160 }, { "grad_norm": 0.09387141466140747, "learning_rate": 1.715948947949642e-05, "loss": 0.0099, "step": 74170 }, { "grad_norm": 0.06251483410596848, "learning_rate": 1.714702318818593e-05, "loss": 0.0107, "step": 74180 }, { "grad_norm": 0.14361488819122314, "learning_rate": 1.7134560489627592e-05, "loss": 0.0123, "step": 74190 }, { "grad_norm": 0.41594621539115906, "learning_rate": 1.712210138518431e-05, "loss": 0.012, "step": 74200 }, { "grad_norm": 0.18172666430473328, "learning_rate": 1.71096458762186e-05, "loss": 0.0182, "step": 74210 }, { "grad_norm": 0.3217287063598633, "learning_rate": 1.7097193964092552e-05, "loss": 0.0169, "step": 74220 }, { "grad_norm": 0.231314018368721, "learning_rate": 1.708474565016792e-05, "loss": 0.0154, "step": 74230 }, { "grad_norm": 0.09900879114866257, "learning_rate": 1.7072300935806e-05, "loss": 0.0081, "step": 74240 }, { "grad_norm": 0.3199383020401001, "learning_rate": 1.705985982236776e-05, "loss": 0.015, "step": 74250 }, { "grad_norm": 0.17412015795707703, "learning_rate": 1.7047422311213697e-05, "loss": 0.0089, "step": 74260 }, { "grad_norm": 0.1116720661520958, "learning_rate": 1.703498840370401e-05, "loss": 0.0138, "step": 74270 }, { "grad_norm": 0.11757057905197144, "learning_rate": 1.702255810119842e-05, "loss": 0.0114, "step": 74280 }, { "grad_norm": 0.14249743521213531, "learning_rate": 1.7010131405056305e-05, "loss": 0.0095, "step": 74290 }, { "grad_norm": 0.08340033888816833, "learning_rate": 1.6997708316636613e-05, "loss": 0.0106, "step": 74300 }, { "grad_norm": 0.02101672813296318, "learning_rate": 1.6985288837297932e-05, "loss": 0.0165, "step": 74310 }, { "grad_norm": 0.1996362954378128, "learning_rate": 1.6972872968398394e-05, "loss": 0.0123, "step": 74320 }, { "grad_norm": 0.24855934083461761, "learning_rate": 1.6960460711295845e-05, "loss": 0.0113, "step": 74330 }, { "grad_norm": 0.10724585503339767, "learning_rate": 1.6948052067347623e-05, "loss": 0.0112, "step": 74340 }, { "grad_norm": 0.1785961389541626, "learning_rate": 1.693564703791075e-05, "loss": 0.0092, "step": 74350 }, { "grad_norm": 0.08940526098012924, "learning_rate": 1.6923245624341793e-05, "loss": 0.0112, "step": 74360 }, { "grad_norm": 0.09869948029518127, "learning_rate": 1.6910847827996962e-05, "loss": 0.0162, "step": 74370 }, { "grad_norm": 0.11074645072221756, "learning_rate": 1.689845365023206e-05, "loss": 0.0182, "step": 74380 }, { "grad_norm": 0.12607496976852417, "learning_rate": 1.6886063092402505e-05, "loss": 0.0165, "step": 74390 }, { "grad_norm": 0.19488434493541718, "learning_rate": 1.6873676155863317e-05, "loss": 0.0189, "step": 74400 }, { "grad_norm": 0.1461535096168518, "learning_rate": 1.6861292841969085e-05, "loss": 0.0161, "step": 74410 }, { "grad_norm": 0.1105087623000145, "learning_rate": 1.684891315207406e-05, "loss": 0.0146, "step": 74420 }, { "grad_norm": 0.30149441957473755, "learning_rate": 1.6836537087532034e-05, "loss": 0.0117, "step": 74430 }, { "grad_norm": 0.12660765647888184, "learning_rate": 1.6824164649696465e-05, "loss": 0.0088, "step": 74440 }, { "grad_norm": 0.18481490015983582, "learning_rate": 1.681179583992034e-05, "loss": 0.0142, "step": 74450 }, { "grad_norm": 0.11451687663793564, "learning_rate": 1.679943065955636e-05, "loss": 0.0117, "step": 74460 }, { "grad_norm": 0.12944163382053375, "learning_rate": 1.678706910995671e-05, "loss": 0.01, "step": 74470 }, { "grad_norm": 0.2128652036190033, "learning_rate": 1.6774711192473262e-05, "loss": 0.0204, "step": 74480 }, { "grad_norm": 0.13578979671001434, "learning_rate": 1.676235690845743e-05, "loss": 0.0088, "step": 74490 }, { "grad_norm": 0.1440359503030777, "learning_rate": 1.6750006259260285e-05, "loss": 0.0101, "step": 74500 }, { "grad_norm": 0.5382875204086304, "learning_rate": 1.673765924623244e-05, "loss": 0.0224, "step": 74510 }, { "grad_norm": 0.30389055609703064, "learning_rate": 1.67253158707242e-05, "loss": 0.0129, "step": 74520 }, { "grad_norm": 0.15740904211997986, "learning_rate": 1.671297613408537e-05, "loss": 0.0099, "step": 74530 }, { "grad_norm": 0.3357403874397278, "learning_rate": 1.6700640037665444e-05, "loss": 0.0132, "step": 74540 }, { "grad_norm": 0.11735960841178894, "learning_rate": 1.668830758281344e-05, "loss": 0.0119, "step": 74550 }, { "grad_norm": 0.17164501547813416, "learning_rate": 1.6675978770878052e-05, "loss": 0.0133, "step": 74560 }, { "grad_norm": 0.21966063976287842, "learning_rate": 1.6663653603207506e-05, "loss": 0.0154, "step": 74570 }, { "grad_norm": 0.3081835210323334, "learning_rate": 1.6651332081149685e-05, "loss": 0.0151, "step": 74580 }, { "grad_norm": 0.10483203083276749, "learning_rate": 1.6639014206052046e-05, "loss": 0.0162, "step": 74590 }, { "grad_norm": 0.3555958569049835, "learning_rate": 1.6626699979261657e-05, "loss": 0.0107, "step": 74600 }, { "grad_norm": 0.2885671555995941, "learning_rate": 1.6614389402125197e-05, "loss": 0.0076, "step": 74610 }, { "grad_norm": 0.2880932092666626, "learning_rate": 1.66020824759889e-05, "loss": 0.0155, "step": 74620 }, { "grad_norm": 0.27969855070114136, "learning_rate": 1.6589779202198675e-05, "loss": 0.0169, "step": 74630 }, { "grad_norm": 0.24813714623451233, "learning_rate": 1.6577479582099924e-05, "loss": 0.016, "step": 74640 }, { "grad_norm": 0.1649908572435379, "learning_rate": 1.6565183617037788e-05, "loss": 0.0187, "step": 74650 }, { "grad_norm": 0.22183416783809662, "learning_rate": 1.6552891308356894e-05, "loss": 0.0118, "step": 74660 }, { "grad_norm": 0.19292035698890686, "learning_rate": 1.6540602657401526e-05, "loss": 0.0167, "step": 74670 }, { "grad_norm": 0.18581008911132812, "learning_rate": 1.6528317665515537e-05, "loss": 0.0246, "step": 74680 }, { "grad_norm": 0.2171505093574524, "learning_rate": 1.6516036334042416e-05, "loss": 0.0193, "step": 74690 }, { "grad_norm": 0.08059675246477127, "learning_rate": 1.650375866432519e-05, "loss": 0.0148, "step": 74700 }, { "grad_norm": 0.13836443424224854, "learning_rate": 1.649148465770659e-05, "loss": 0.0188, "step": 74710 }, { "grad_norm": 0.13499794900417328, "learning_rate": 1.647921431552883e-05, "loss": 0.0085, "step": 74720 }, { "grad_norm": 0.2580966651439667, "learning_rate": 1.646694763913381e-05, "loss": 0.0178, "step": 74730 }, { "grad_norm": 0.31029024720191956, "learning_rate": 1.6454684629862966e-05, "loss": 0.0127, "step": 74740 }, { "grad_norm": 0.14293727278709412, "learning_rate": 1.6442425289057388e-05, "loss": 0.0133, "step": 74750 }, { "grad_norm": 0.2444761097431183, "learning_rate": 1.6430169618057718e-05, "loss": 0.0138, "step": 74760 }, { "grad_norm": 0.15508492290973663, "learning_rate": 1.641791761820422e-05, "loss": 0.0136, "step": 74770 }, { "grad_norm": 0.38147422671318054, "learning_rate": 1.640566929083676e-05, "loss": 0.0147, "step": 74780 }, { "grad_norm": 0.23581260442733765, "learning_rate": 1.6393424637294823e-05, "loss": 0.0207, "step": 74790 }, { "grad_norm": 0.3590267598628998, "learning_rate": 1.638118365891742e-05, "loss": 0.0111, "step": 74800 }, { "grad_norm": 0.19721141457557678, "learning_rate": 1.636894635704323e-05, "loss": 0.0238, "step": 74810 }, { "grad_norm": 0.23081497848033905, "learning_rate": 1.6356712733010517e-05, "loss": 0.0131, "step": 74820 }, { "grad_norm": 0.15211205184459686, "learning_rate": 1.6344482788157096e-05, "loss": 0.0104, "step": 74830 }, { "grad_norm": 0.06432944536209106, "learning_rate": 1.6332256523820444e-05, "loss": 0.0097, "step": 74840 }, { "grad_norm": 0.2339099943637848, "learning_rate": 1.6320033941337597e-05, "loss": 0.0137, "step": 74850 }, { "grad_norm": 0.13662056624889374, "learning_rate": 1.6307815042045216e-05, "loss": 0.0138, "step": 74860 }, { "grad_norm": 0.18328581750392914, "learning_rate": 1.6295599827279513e-05, "loss": 0.013, "step": 74870 }, { "grad_norm": 0.10798414051532745, "learning_rate": 1.6283388298376356e-05, "loss": 0.0136, "step": 74880 }, { "grad_norm": 0.07291572540998459, "learning_rate": 1.6271180456671125e-05, "loss": 0.0136, "step": 74890 }, { "grad_norm": 0.13931505382061005, "learning_rate": 1.625897630349893e-05, "loss": 0.0096, "step": 74900 }, { "grad_norm": 0.13912418484687805, "learning_rate": 1.6246775840194335e-05, "loss": 0.0105, "step": 74910 }, { "grad_norm": 0.231278195977211, "learning_rate": 1.6234579068091604e-05, "loss": 0.0139, "step": 74920 }, { "grad_norm": 0.19933712482452393, "learning_rate": 1.6222385988524524e-05, "loss": 0.0162, "step": 74930 }, { "grad_norm": 0.24649663269519806, "learning_rate": 1.6210196602826544e-05, "loss": 0.018, "step": 74940 }, { "grad_norm": 0.21217671036720276, "learning_rate": 1.6198010912330646e-05, "loss": 0.0126, "step": 74950 }, { "grad_norm": 0.31912317872047424, "learning_rate": 1.6185828918369456e-05, "loss": 0.0243, "step": 74960 }, { "grad_norm": 0.1492684781551361, "learning_rate": 1.617365062227517e-05, "loss": 0.0102, "step": 74970 }, { "grad_norm": 0.29027417302131653, "learning_rate": 1.616147602537962e-05, "loss": 0.0115, "step": 74980 }, { "grad_norm": 0.039072729647159576, "learning_rate": 1.614930512901415e-05, "loss": 0.011, "step": 74990 }, { "grad_norm": 0.14569038152694702, "learning_rate": 1.6137137934509804e-05, "loss": 0.0108, "step": 75000 }, { "grad_norm": 0.30495354533195496, "learning_rate": 1.6124974443197123e-05, "loss": 0.0119, "step": 75010 }, { "grad_norm": 0.0759362131357193, "learning_rate": 1.6112814656406306e-05, "loss": 0.0118, "step": 75020 }, { "grad_norm": 0.10596752166748047, "learning_rate": 1.6100658575467125e-05, "loss": 0.0088, "step": 75030 }, { "grad_norm": 0.23828372359275818, "learning_rate": 1.6088506201708974e-05, "loss": 0.0143, "step": 75040 }, { "grad_norm": 0.2606778144836426, "learning_rate": 1.6076357536460785e-05, "loss": 0.0111, "step": 75050 }, { "grad_norm": 0.30373281240463257, "learning_rate": 1.606421258105113e-05, "loss": 0.0141, "step": 75060 }, { "grad_norm": 0.0889582708477974, "learning_rate": 1.6052071336808183e-05, "loss": 0.0149, "step": 75070 }, { "grad_norm": 0.30367374420166016, "learning_rate": 1.6039933805059658e-05, "loss": 0.0097, "step": 75080 }, { "grad_norm": 0.362496942281723, "learning_rate": 1.6027799987132915e-05, "loss": 0.0101, "step": 75090 }, { "grad_norm": 0.3759719729423523, "learning_rate": 1.6015669884354888e-05, "loss": 0.0177, "step": 75100 }, { "grad_norm": 0.16153450310230255, "learning_rate": 1.6003543498052125e-05, "loss": 0.0176, "step": 75110 }, { "grad_norm": 0.16718335449695587, "learning_rate": 1.5991420829550715e-05, "loss": 0.0169, "step": 75120 }, { "grad_norm": 0.2670115828514099, "learning_rate": 1.5979301880176405e-05, "loss": 0.0138, "step": 75130 }, { "grad_norm": 0.16265493631362915, "learning_rate": 1.5967186651254485e-05, "loss": 0.0134, "step": 75140 }, { "grad_norm": 0.21714308857917786, "learning_rate": 1.5955075144109853e-05, "loss": 0.0162, "step": 75150 }, { "grad_norm": 0.1597878783941269, "learning_rate": 1.594296736006703e-05, "loss": 0.0206, "step": 75160 }, { "grad_norm": 0.0717330202460289, "learning_rate": 1.59308633004501e-05, "loss": 0.0173, "step": 75170 }, { "grad_norm": 0.07783160358667374, "learning_rate": 1.5918762966582724e-05, "loss": 0.0108, "step": 75180 }, { "grad_norm": 0.05192325636744499, "learning_rate": 1.5906666359788203e-05, "loss": 0.0122, "step": 75190 }, { "grad_norm": 0.392317533493042, "learning_rate": 1.589457348138938e-05, "loss": 0.0083, "step": 75200 }, { "grad_norm": 0.23319090902805328, "learning_rate": 1.588248433270872e-05, "loss": 0.0144, "step": 75210 }, { "grad_norm": 0.11817579716444016, "learning_rate": 1.587039891506828e-05, "loss": 0.016, "step": 75220 }, { "grad_norm": 0.12461983412504196, "learning_rate": 1.5858317229789715e-05, "loss": 0.01, "step": 75230 }, { "grad_norm": 0.07235903292894363, "learning_rate": 1.584623927819423e-05, "loss": 0.0118, "step": 75240 }, { "grad_norm": 0.2538689076900482, "learning_rate": 1.583416506160269e-05, "loss": 0.01, "step": 75250 }, { "grad_norm": 0.34129849076271057, "learning_rate": 1.5822094581335472e-05, "loss": 0.0149, "step": 75260 }, { "grad_norm": 0.13284780085086823, "learning_rate": 1.5810027838712604e-05, "loss": 0.0192, "step": 75270 }, { "grad_norm": 0.08666323870420456, "learning_rate": 1.5797964835053687e-05, "loss": 0.0101, "step": 75280 }, { "grad_norm": 0.08123563230037689, "learning_rate": 1.5785905571677933e-05, "loss": 0.0107, "step": 75290 }, { "grad_norm": 0.2754132151603699, "learning_rate": 1.5773850049904086e-05, "loss": 0.0099, "step": 75300 }, { "grad_norm": 0.12007086724042892, "learning_rate": 1.5761798271050543e-05, "loss": 0.0099, "step": 75310 }, { "grad_norm": 0.1812044084072113, "learning_rate": 1.5749750236435277e-05, "loss": 0.0092, "step": 75320 }, { "grad_norm": 0.3422785997390747, "learning_rate": 1.5737705947375815e-05, "loss": 0.0125, "step": 75330 }, { "grad_norm": 0.1533346027135849, "learning_rate": 1.5725665405189322e-05, "loss": 0.0145, "step": 75340 }, { "grad_norm": 0.2695179283618927, "learning_rate": 1.571362861119253e-05, "loss": 0.0102, "step": 75350 }, { "grad_norm": 0.378608375787735, "learning_rate": 1.570159556670177e-05, "loss": 0.0116, "step": 75360 }, { "grad_norm": 0.1320800483226776, "learning_rate": 1.5689566273032946e-05, "loss": 0.0101, "step": 75370 }, { "grad_norm": 0.14113914966583252, "learning_rate": 1.5677540731501583e-05, "loss": 0.0197, "step": 75380 }, { "grad_norm": 0.3408929705619812, "learning_rate": 1.5665518943422746e-05, "loss": 0.0168, "step": 75390 }, { "grad_norm": 0.16116845607757568, "learning_rate": 1.5653500910111134e-05, "loss": 0.0174, "step": 75400 }, { "grad_norm": 0.24056671559810638, "learning_rate": 1.5641486632881023e-05, "loss": 0.0143, "step": 75410 }, { "grad_norm": 0.1792442351579666, "learning_rate": 1.56294761130463e-05, "loss": 0.0147, "step": 75420 }, { "grad_norm": 0.20628215372562408, "learning_rate": 1.5617469351920373e-05, "loss": 0.0112, "step": 75430 }, { "grad_norm": 0.1268717646598816, "learning_rate": 1.5605466350816315e-05, "loss": 0.0142, "step": 75440 }, { "grad_norm": 0.2518976330757141, "learning_rate": 1.559346711104674e-05, "loss": 0.014, "step": 75450 }, { "grad_norm": 0.25657203793525696, "learning_rate": 1.5581471633923883e-05, "loss": 0.0081, "step": 75460 }, { "grad_norm": 0.24184229969978333, "learning_rate": 1.5569479920759505e-05, "loss": 0.0141, "step": 75470 }, { "grad_norm": 0.0726408138871193, "learning_rate": 1.5557491972865072e-05, "loss": 0.0068, "step": 75480 }, { "grad_norm": 0.20137909054756165, "learning_rate": 1.554550779155152e-05, "loss": 0.0101, "step": 75490 }, { "grad_norm": 0.30015626549720764, "learning_rate": 1.5533527378129453e-05, "loss": 0.0123, "step": 75500 }, { "grad_norm": 0.11035211384296417, "learning_rate": 1.552155073390899e-05, "loss": 0.0128, "step": 75510 }, { "grad_norm": 0.049947794526815414, "learning_rate": 1.5509577860199903e-05, "loss": 0.0241, "step": 75520 }, { "grad_norm": 0.18551839888095856, "learning_rate": 1.5497608758311525e-05, "loss": 0.0101, "step": 75530 }, { "grad_norm": 0.23158013820648193, "learning_rate": 1.5485643429552797e-05, "loss": 0.0158, "step": 75540 }, { "grad_norm": 0.0814993605017662, "learning_rate": 1.5473681875232197e-05, "loss": 0.0113, "step": 75550 }, { "grad_norm": 0.1215275451540947, "learning_rate": 1.5461724096657837e-05, "loss": 0.0101, "step": 75560 }, { "grad_norm": 0.1343095898628235, "learning_rate": 1.5449770095137416e-05, "loss": 0.0123, "step": 75570 }, { "grad_norm": 0.10324207693338394, "learning_rate": 1.5437819871978178e-05, "loss": 0.0133, "step": 75580 }, { "grad_norm": 0.1191205382347107, "learning_rate": 1.5425873428486992e-05, "loss": 0.0185, "step": 75590 }, { "grad_norm": 0.11041747778654099, "learning_rate": 1.5413930765970312e-05, "loss": 0.0134, "step": 75600 }, { "grad_norm": 0.07323943078517914, "learning_rate": 1.5401991885734175e-05, "loss": 0.0147, "step": 75610 }, { "grad_norm": 0.3522776961326599, "learning_rate": 1.5390056789084174e-05, "loss": 0.0128, "step": 75620 }, { "grad_norm": 0.12160462886095047, "learning_rate": 1.5378125477325545e-05, "loss": 0.0125, "step": 75630 }, { "grad_norm": 0.2676170766353607, "learning_rate": 1.5366197951763046e-05, "loss": 0.0173, "step": 75640 }, { "grad_norm": 0.19600196182727814, "learning_rate": 1.5354274213701075e-05, "loss": 0.0151, "step": 75650 }, { "grad_norm": 0.06207134947180748, "learning_rate": 1.5342354264443566e-05, "loss": 0.0081, "step": 75660 }, { "grad_norm": 0.3672899901866913, "learning_rate": 1.533043810529411e-05, "loss": 0.0166, "step": 75670 }, { "grad_norm": 0.1435718685388565, "learning_rate": 1.5318525737555807e-05, "loss": 0.019, "step": 75680 }, { "grad_norm": 0.11810187995433807, "learning_rate": 1.53066171625314e-05, "loss": 0.018, "step": 75690 }, { "grad_norm": 0.09645764529705048, "learning_rate": 1.529471238152317e-05, "loss": 0.0172, "step": 75700 }, { "grad_norm": 0.08797244727611542, "learning_rate": 1.5282811395833023e-05, "loss": 0.0091, "step": 75710 }, { "grad_norm": 0.10872778296470642, "learning_rate": 1.5270914206762397e-05, "loss": 0.0141, "step": 75720 }, { "grad_norm": 0.34369856119155884, "learning_rate": 1.525902081561241e-05, "loss": 0.0119, "step": 75730 }, { "grad_norm": 0.24302269518375397, "learning_rate": 1.5247131223683664e-05, "loss": 0.0096, "step": 75740 }, { "grad_norm": 0.22800008952617645, "learning_rate": 1.5235245432276408e-05, "loss": 0.0136, "step": 75750 }, { "grad_norm": 0.12345710396766663, "learning_rate": 1.5223363442690424e-05, "loss": 0.011, "step": 75760 }, { "grad_norm": 0.4310462474822998, "learning_rate": 1.5211485256225133e-05, "loss": 0.0123, "step": 75770 }, { "grad_norm": 0.12677216529846191, "learning_rate": 1.519961087417951e-05, "loss": 0.0066, "step": 75780 }, { "grad_norm": 0.19253133237361908, "learning_rate": 1.5187740297852137e-05, "loss": 0.0111, "step": 75790 }, { "grad_norm": 0.09796358644962311, "learning_rate": 1.517587352854113e-05, "loss": 0.016, "step": 75800 }, { "grad_norm": 0.31955912709236145, "learning_rate": 1.516401056754423e-05, "loss": 0.0129, "step": 75810 }, { "grad_norm": 0.2907734811306, "learning_rate": 1.5152151416158783e-05, "loss": 0.0148, "step": 75820 }, { "grad_norm": 0.14191162586212158, "learning_rate": 1.5140296075681642e-05, "loss": 0.0149, "step": 75830 }, { "grad_norm": 0.21064594388008118, "learning_rate": 1.5128444547409331e-05, "loss": 0.0108, "step": 75840 }, { "grad_norm": 0.1677151918411255, "learning_rate": 1.5116596832637864e-05, "loss": 0.0113, "step": 75850 }, { "grad_norm": 0.2942230701446533, "learning_rate": 1.5104752932662952e-05, "loss": 0.0141, "step": 75860 }, { "grad_norm": 0.2940174341201782, "learning_rate": 1.5092912848779778e-05, "loss": 0.0114, "step": 75870 }, { "grad_norm": 0.17331910133361816, "learning_rate": 1.5081076582283194e-05, "loss": 0.0124, "step": 75880 }, { "grad_norm": 0.16559453308582306, "learning_rate": 1.5069244134467553e-05, "loss": 0.0099, "step": 75890 }, { "grad_norm": 0.12401477247476578, "learning_rate": 1.5057415506626876e-05, "loss": 0.0128, "step": 75900 }, { "grad_norm": 0.1389598399400711, "learning_rate": 1.5045590700054668e-05, "loss": 0.0101, "step": 75910 }, { "grad_norm": 0.24421165883541107, "learning_rate": 1.5033769716044144e-05, "loss": 0.0157, "step": 75920 }, { "grad_norm": 0.5327466726303101, "learning_rate": 1.5021952555887975e-05, "loss": 0.0124, "step": 75930 }, { "grad_norm": 0.11509465426206589, "learning_rate": 1.50101392208785e-05, "loss": 0.0094, "step": 75940 }, { "grad_norm": 0.09421755373477936, "learning_rate": 1.4998329712307579e-05, "loss": 0.0113, "step": 75950 }, { "grad_norm": 0.2117883712053299, "learning_rate": 1.4986524031466708e-05, "loss": 0.0105, "step": 75960 }, { "grad_norm": 0.2581215798854828, "learning_rate": 1.4974722179646916e-05, "loss": 0.0103, "step": 75970 }, { "grad_norm": 0.272634357213974, "learning_rate": 1.4962924158138847e-05, "loss": 0.0169, "step": 75980 }, { "grad_norm": 0.0926639661192894, "learning_rate": 1.4951129968232714e-05, "loss": 0.0085, "step": 75990 }, { "grad_norm": 0.13208052515983582, "learning_rate": 1.4939339611218329e-05, "loss": 0.0101, "step": 76000 }, { "grad_norm": 0.19846269488334656, "learning_rate": 1.4927553088385032e-05, "loss": 0.016, "step": 76010 }, { "grad_norm": 0.03929700329899788, "learning_rate": 1.4915770401021801e-05, "loss": 0.0099, "step": 76020 }, { "grad_norm": 0.2027001529932022, "learning_rate": 1.4903991550417185e-05, "loss": 0.0093, "step": 76030 }, { "grad_norm": 0.22350294888019562, "learning_rate": 1.4892216537859277e-05, "loss": 0.014, "step": 76040 }, { "grad_norm": 0.24012891948223114, "learning_rate": 1.4880445364635775e-05, "loss": 0.0139, "step": 76050 }, { "grad_norm": 0.168310284614563, "learning_rate": 1.4868678032033973e-05, "loss": 0.0158, "step": 76060 }, { "grad_norm": 0.16901379823684692, "learning_rate": 1.4856914541340739e-05, "loss": 0.0111, "step": 76070 }, { "grad_norm": 0.18776468932628632, "learning_rate": 1.4845154893842472e-05, "loss": 0.0206, "step": 76080 }, { "grad_norm": 0.3517588675022125, "learning_rate": 1.4833399090825234e-05, "loss": 0.0241, "step": 76090 }, { "grad_norm": 0.14582134783267975, "learning_rate": 1.4821647133574562e-05, "loss": 0.0123, "step": 76100 }, { "grad_norm": 0.09593373537063599, "learning_rate": 1.4809899023375706e-05, "loss": 0.0133, "step": 76110 }, { "grad_norm": 0.22416764497756958, "learning_rate": 1.4798154761513366e-05, "loss": 0.0125, "step": 76120 }, { "grad_norm": 0.08101070672273636, "learning_rate": 1.4786414349271916e-05, "loss": 0.0084, "step": 76130 }, { "grad_norm": 0.03720562905073166, "learning_rate": 1.4774677787935226e-05, "loss": 0.0152, "step": 76140 }, { "grad_norm": 0.20836995542049408, "learning_rate": 1.4762945078786838e-05, "loss": 0.0115, "step": 76150 }, { "grad_norm": 0.2704067528247833, "learning_rate": 1.4751216223109777e-05, "loss": 0.0127, "step": 76160 }, { "grad_norm": 0.14448828995227814, "learning_rate": 1.4739491222186714e-05, "loss": 0.0141, "step": 76170 }, { "grad_norm": 0.18757690489292145, "learning_rate": 1.4727770077299879e-05, "loss": 0.0145, "step": 76180 }, { "grad_norm": 0.07930966466665268, "learning_rate": 1.4716052789731094e-05, "loss": 0.0102, "step": 76190 }, { "grad_norm": 0.13020287454128265, "learning_rate": 1.470433936076171e-05, "loss": 0.019, "step": 76200 }, { "grad_norm": 0.07741334289312363, "learning_rate": 1.4692629791672719e-05, "loss": 0.0114, "step": 76210 }, { "grad_norm": 0.05044851824641228, "learning_rate": 1.468092408374464e-05, "loss": 0.0095, "step": 76220 }, { "grad_norm": 0.14073480665683746, "learning_rate": 1.4669222238257602e-05, "loss": 0.0131, "step": 76230 }, { "grad_norm": 0.1633066087961197, "learning_rate": 1.4657524256491301e-05, "loss": 0.0121, "step": 76240 }, { "grad_norm": 0.3323080241680145, "learning_rate": 1.464583013972502e-05, "loss": 0.0125, "step": 76250 }, { "grad_norm": 0.20399664342403412, "learning_rate": 1.4634139889237591e-05, "loss": 0.0117, "step": 76260 }, { "grad_norm": 0.16729524731636047, "learning_rate": 1.4622453506307448e-05, "loss": 0.0088, "step": 76270 }, { "grad_norm": 0.09631136804819107, "learning_rate": 1.461077099221262e-05, "loss": 0.0078, "step": 76280 }, { "grad_norm": 0.03720160946249962, "learning_rate": 1.4599092348230652e-05, "loss": 0.0077, "step": 76290 }, { "grad_norm": 0.1072956994175911, "learning_rate": 1.4587417575638724e-05, "loss": 0.0113, "step": 76300 }, { "grad_norm": 0.20180360972881317, "learning_rate": 1.4575746675713563e-05, "loss": 0.0134, "step": 76310 }, { "grad_norm": 0.106438048183918, "learning_rate": 1.4564079649731504e-05, "loss": 0.0105, "step": 76320 }, { "grad_norm": 0.20369097590446472, "learning_rate": 1.4552416498968402e-05, "loss": 0.0081, "step": 76330 }, { "grad_norm": 0.11031761765480042, "learning_rate": 1.4540757224699747e-05, "loss": 0.0128, "step": 76340 }, { "grad_norm": 0.10891687870025635, "learning_rate": 1.452910182820056e-05, "loss": 0.0134, "step": 76350 }, { "grad_norm": 0.14170922338962555, "learning_rate": 1.4517450310745468e-05, "loss": 0.0141, "step": 76360 }, { "grad_norm": 0.3117956817150116, "learning_rate": 1.450580267360866e-05, "loss": 0.0143, "step": 76370 }, { "grad_norm": 0.11366190016269684, "learning_rate": 1.4494158918063922e-05, "loss": 0.0057, "step": 76380 }, { "grad_norm": 0.23012609779834747, "learning_rate": 1.4482519045384568e-05, "loss": 0.0104, "step": 76390 }, { "grad_norm": 0.04126347601413727, "learning_rate": 1.4470883056843548e-05, "loss": 0.0127, "step": 76400 }, { "grad_norm": 0.19622401893138885, "learning_rate": 1.4459250953713322e-05, "loss": 0.017, "step": 76410 }, { "grad_norm": 0.09809926152229309, "learning_rate": 1.444762273726598e-05, "loss": 0.0162, "step": 76420 }, { "grad_norm": 0.09662871062755585, "learning_rate": 1.443599840877316e-05, "loss": 0.0125, "step": 76430 }, { "grad_norm": 0.12200409173965454, "learning_rate": 1.4424377969506103e-05, "loss": 0.0118, "step": 76440 }, { "grad_norm": 1.1268669366836548, "learning_rate": 1.4412761420735565e-05, "loss": 0.0152, "step": 76450 }, { "grad_norm": 0.1193406954407692, "learning_rate": 1.4401148763731953e-05, "loss": 0.0203, "step": 76460 }, { "grad_norm": 0.15361803770065308, "learning_rate": 1.4389539999765173e-05, "loss": 0.0083, "step": 76470 }, { "grad_norm": 0.25302204489707947, "learning_rate": 1.4377935130104758e-05, "loss": 0.0095, "step": 76480 }, { "grad_norm": 0.08261139690876007, "learning_rate": 1.4366334156019801e-05, "loss": 0.0111, "step": 76490 }, { "grad_norm": 0.20135487616062164, "learning_rate": 1.435473707877898e-05, "loss": 0.0137, "step": 76500 }, { "grad_norm": 0.5856080055236816, "learning_rate": 1.4343143899650508e-05, "loss": 0.0202, "step": 76510 }, { "grad_norm": 0.31966468691825867, "learning_rate": 1.4331554619902204e-05, "loss": 0.0137, "step": 76520 }, { "grad_norm": 0.3610597848892212, "learning_rate": 1.4319969240801484e-05, "loss": 0.0119, "step": 76530 }, { "grad_norm": 0.10028965771198273, "learning_rate": 1.430838776361526e-05, "loss": 0.013, "step": 76540 }, { "grad_norm": 0.13150908052921295, "learning_rate": 1.4296810189610093e-05, "loss": 0.0113, "step": 76550 }, { "grad_norm": 0.11099347472190857, "learning_rate": 1.4285236520052087e-05, "loss": 0.0074, "step": 76560 }, { "grad_norm": 0.11127328872680664, "learning_rate": 1.4273666756206933e-05, "loss": 0.0096, "step": 76570 }, { "grad_norm": 0.160757914185524, "learning_rate": 1.4262100899339848e-05, "loss": 0.0131, "step": 76580 }, { "grad_norm": 0.3130905330181122, "learning_rate": 1.4250538950715696e-05, "loss": 0.0131, "step": 76590 }, { "grad_norm": 0.13202990591526031, "learning_rate": 1.4238980911598842e-05, "loss": 0.0126, "step": 76600 }, { "grad_norm": 0.357383668422699, "learning_rate": 1.4227426783253267e-05, "loss": 0.0161, "step": 76610 }, { "grad_norm": 0.2944970726966858, "learning_rate": 1.4215876566942521e-05, "loss": 0.0132, "step": 76620 }, { "grad_norm": 0.12959660589694977, "learning_rate": 1.4204330263929721e-05, "loss": 0.0079, "step": 76630 }, { "grad_norm": 0.18541809916496277, "learning_rate": 1.4192787875477537e-05, "loss": 0.016, "step": 76640 }, { "grad_norm": 0.3510717451572418, "learning_rate": 1.4181249402848246e-05, "loss": 0.0132, "step": 76650 }, { "grad_norm": 0.11294475942850113, "learning_rate": 1.4169714847303656e-05, "loss": 0.0093, "step": 76660 }, { "grad_norm": 0.34598270058631897, "learning_rate": 1.4158184210105197e-05, "loss": 0.0149, "step": 76670 }, { "grad_norm": 0.08231242001056671, "learning_rate": 1.4146657492513787e-05, "loss": 0.0117, "step": 76680 }, { "grad_norm": 0.09087737649679184, "learning_rate": 1.4135134695790047e-05, "loss": 0.0079, "step": 76690 }, { "grad_norm": 0.20866435766220093, "learning_rate": 1.4123615821194037e-05, "loss": 0.0093, "step": 76700 }, { "grad_norm": 0.11761024594306946, "learning_rate": 1.4112100869985473e-05, "loss": 0.0149, "step": 76710 }, { "grad_norm": 0.16900594532489777, "learning_rate": 1.4100589843423589e-05, "loss": 0.0158, "step": 76720 }, { "grad_norm": 0.1634371280670166, "learning_rate": 1.408908274276723e-05, "loss": 0.0076, "step": 76730 }, { "grad_norm": 0.2903124690055847, "learning_rate": 1.4077579569274786e-05, "loss": 0.0158, "step": 76740 }, { "grad_norm": 0.12648224830627441, "learning_rate": 1.4066080324204245e-05, "loss": 0.0083, "step": 76750 }, { "grad_norm": 0.20375601947307587, "learning_rate": 1.405458500881312e-05, "loss": 0.0247, "step": 76760 }, { "grad_norm": 0.16769297420978546, "learning_rate": 1.4043093624358538e-05, "loss": 0.0112, "step": 76770 }, { "grad_norm": 0.14089341461658478, "learning_rate": 1.4031606172097189e-05, "loss": 0.0166, "step": 76780 }, { "grad_norm": 0.09771765023469925, "learning_rate": 1.40201226532853e-05, "loss": 0.0094, "step": 76790 }, { "grad_norm": 0.10475598275661469, "learning_rate": 1.4008643069178701e-05, "loss": 0.0192, "step": 76800 }, { "grad_norm": 0.1150016263127327, "learning_rate": 1.399716742103278e-05, "loss": 0.0106, "step": 76810 }, { "grad_norm": 0.12406332045793533, "learning_rate": 1.398569571010252e-05, "loss": 0.0088, "step": 76820 }, { "grad_norm": 0.15226013958454132, "learning_rate": 1.3974227937642414e-05, "loss": 0.0119, "step": 76830 }, { "grad_norm": 0.12137465924024582, "learning_rate": 1.3962764104906595e-05, "loss": 0.0079, "step": 76840 }, { "grad_norm": 0.18718309700489044, "learning_rate": 1.3951304213148696e-05, "loss": 0.014, "step": 76850 }, { "grad_norm": 0.07352142035961151, "learning_rate": 1.3939848263621979e-05, "loss": 0.0106, "step": 76860 }, { "grad_norm": 0.28389570116996765, "learning_rate": 1.3928396257579219e-05, "loss": 0.0155, "step": 76870 }, { "grad_norm": 0.15142183005809784, "learning_rate": 1.3916948196272833e-05, "loss": 0.0084, "step": 76880 }, { "grad_norm": 0.06778094172477722, "learning_rate": 1.3905504080954734e-05, "loss": 0.0167, "step": 76890 }, { "grad_norm": 0.20662184059619904, "learning_rate": 1.3894063912876454e-05, "loss": 0.0104, "step": 76900 }, { "grad_norm": 0.1541977822780609, "learning_rate": 1.3882627693289046e-05, "loss": 0.0115, "step": 76910 }, { "grad_norm": 0.07293678820133209, "learning_rate": 1.3871195423443183e-05, "loss": 0.0134, "step": 76920 }, { "grad_norm": 0.2567299008369446, "learning_rate": 1.3859767104589044e-05, "loss": 0.0106, "step": 76930 }, { "grad_norm": 0.14731749892234802, "learning_rate": 1.3848342737976467e-05, "loss": 0.008, "step": 76940 }, { "grad_norm": 0.24236823618412018, "learning_rate": 1.3836922324854757e-05, "loss": 0.0163, "step": 76950 }, { "grad_norm": 0.2961055338382721, "learning_rate": 1.382550586647287e-05, "loss": 0.021, "step": 76960 }, { "grad_norm": 0.07633163779973984, "learning_rate": 1.3814093364079257e-05, "loss": 0.0112, "step": 76970 }, { "grad_norm": 0.1298360973596573, "learning_rate": 1.3802684818921991e-05, "loss": 0.0186, "step": 76980 }, { "grad_norm": 0.0919773131608963, "learning_rate": 1.3791280232248687e-05, "loss": 0.0135, "step": 76990 }, { "grad_norm": 0.11458154767751694, "learning_rate": 1.3779879605306555e-05, "loss": 0.0148, "step": 77000 }, { "grad_norm": 0.19568441808223724, "learning_rate": 1.376848293934232e-05, "loss": 0.0115, "step": 77010 }, { "grad_norm": 0.22500203549861908, "learning_rate": 1.3757090235602316e-05, "loss": 0.0111, "step": 77020 }, { "grad_norm": 0.20736342668533325, "learning_rate": 1.3745701495332447e-05, "loss": 0.0156, "step": 77030 }, { "grad_norm": 0.17928071320056915, "learning_rate": 1.3734316719778145e-05, "loss": 0.0108, "step": 77040 }, { "grad_norm": 0.18690089881420135, "learning_rate": 1.3722935910184454e-05, "loss": 0.0113, "step": 77050 }, { "grad_norm": 0.1819470375776291, "learning_rate": 1.3711559067795921e-05, "loss": 0.0156, "step": 77060 }, { "grad_norm": 0.3200500011444092, "learning_rate": 1.3700186193856756e-05, "loss": 0.0118, "step": 77070 }, { "grad_norm": 0.3480775058269501, "learning_rate": 1.368881728961064e-05, "loss": 0.0126, "step": 77080 }, { "grad_norm": 0.11490116268396378, "learning_rate": 1.3677452356300885e-05, "loss": 0.0093, "step": 77090 }, { "grad_norm": 0.1141878142952919, "learning_rate": 1.3666091395170321e-05, "loss": 0.0092, "step": 77100 }, { "grad_norm": 0.0815487876534462, "learning_rate": 1.3654734407461384e-05, "loss": 0.0128, "step": 77110 }, { "grad_norm": 0.3006700575351715, "learning_rate": 1.3643381394416022e-05, "loss": 0.0143, "step": 77120 }, { "grad_norm": 0.12671709060668945, "learning_rate": 1.3632032357275838e-05, "loss": 0.0195, "step": 77130 }, { "grad_norm": 0.27306780219078064, "learning_rate": 1.3620687297281898e-05, "loss": 0.0112, "step": 77140 }, { "grad_norm": 0.07421587407588959, "learning_rate": 1.3609346215674917e-05, "loss": 0.0102, "step": 77150 }, { "grad_norm": 0.14018100500106812, "learning_rate": 1.3598009113695109e-05, "loss": 0.0122, "step": 77160 }, { "grad_norm": 0.05953022092580795, "learning_rate": 1.358667599258231e-05, "loss": 0.017, "step": 77170 }, { "grad_norm": 0.16135331988334656, "learning_rate": 1.357534685357586e-05, "loss": 0.0077, "step": 77180 }, { "grad_norm": 0.1639585942029953, "learning_rate": 1.3564021697914715e-05, "loss": 0.0126, "step": 77190 }, { "grad_norm": 0.15484319627285004, "learning_rate": 1.3552700526837381e-05, "loss": 0.0135, "step": 77200 }, { "grad_norm": 0.1435990035533905, "learning_rate": 1.3541383341581925e-05, "loss": 0.0123, "step": 77210 }, { "grad_norm": 0.1854289174079895, "learning_rate": 1.3530070143385965e-05, "loss": 0.0142, "step": 77220 }, { "grad_norm": 0.10693243891000748, "learning_rate": 1.3518760933486702e-05, "loss": 0.0124, "step": 77230 }, { "grad_norm": 0.3758532702922821, "learning_rate": 1.3507455713120908e-05, "loss": 0.016, "step": 77240 }, { "grad_norm": 0.2955920994281769, "learning_rate": 1.3496154483524876e-05, "loss": 0.0118, "step": 77250 }, { "grad_norm": 0.4797990322113037, "learning_rate": 1.3484857245934506e-05, "loss": 0.0139, "step": 77260 }, { "grad_norm": 0.1446453481912613, "learning_rate": 1.3473564001585248e-05, "loss": 0.0081, "step": 77270 }, { "grad_norm": 0.06443210691213608, "learning_rate": 1.3462274751712129e-05, "loss": 0.011, "step": 77280 }, { "grad_norm": 0.053499672561883926, "learning_rate": 1.3450989497549698e-05, "loss": 0.0079, "step": 77290 }, { "grad_norm": 0.8171511292457581, "learning_rate": 1.3439708240332115e-05, "loss": 0.0315, "step": 77300 }, { "grad_norm": 0.13906565308570862, "learning_rate": 1.3428430981293043e-05, "loss": 0.0125, "step": 77310 }, { "grad_norm": 0.0752766951918602, "learning_rate": 1.3417157721665808e-05, "loss": 0.0119, "step": 77320 }, { "grad_norm": 0.11321926862001419, "learning_rate": 1.3405888462683186e-05, "loss": 0.0106, "step": 77330 }, { "grad_norm": 0.10311386734247208, "learning_rate": 1.3394623205577595e-05, "loss": 0.0088, "step": 77340 }, { "grad_norm": 0.3250442147254944, "learning_rate": 1.3383361951580963e-05, "loss": 0.0208, "step": 77350 }, { "grad_norm": 0.1585930585861206, "learning_rate": 1.337210470192483e-05, "loss": 0.0128, "step": 77360 }, { "grad_norm": 0.228411465883255, "learning_rate": 1.3360851457840245e-05, "loss": 0.0141, "step": 77370 }, { "grad_norm": 0.27410900592803955, "learning_rate": 1.3349602220557856e-05, "loss": 0.0118, "step": 77380 }, { "grad_norm": 0.31487300992012024, "learning_rate": 1.3338356991307865e-05, "loss": 0.0104, "step": 77390 }, { "grad_norm": 0.4924975037574768, "learning_rate": 1.3327115771320048e-05, "loss": 0.0229, "step": 77400 }, { "grad_norm": 0.10056067258119583, "learning_rate": 1.3315878561823697e-05, "loss": 0.0104, "step": 77410 }, { "grad_norm": 0.07522787153720856, "learning_rate": 1.330464536404773e-05, "loss": 0.011, "step": 77420 }, { "grad_norm": 0.23106594383716583, "learning_rate": 1.3293416179220563e-05, "loss": 0.0154, "step": 77430 }, { "grad_norm": 0.2934294044971466, "learning_rate": 1.328219100857021e-05, "loss": 0.0145, "step": 77440 }, { "grad_norm": 0.1288873702287674, "learning_rate": 1.327096985332425e-05, "loss": 0.011, "step": 77450 }, { "grad_norm": 0.07120807468891144, "learning_rate": 1.3259752714709816e-05, "loss": 0.0097, "step": 77460 }, { "grad_norm": 0.05731625109910965, "learning_rate": 1.3248539593953569e-05, "loss": 0.0099, "step": 77470 }, { "grad_norm": 0.1527814269065857, "learning_rate": 1.323733049228178e-05, "loss": 0.0104, "step": 77480 }, { "grad_norm": 0.17850488424301147, "learning_rate": 1.3226125410920276e-05, "loss": 0.0144, "step": 77490 }, { "grad_norm": 0.35695260763168335, "learning_rate": 1.3214924351094388e-05, "loss": 0.017, "step": 77500 }, { "grad_norm": 0.290440171957016, "learning_rate": 1.3203727314029069e-05, "loss": 0.0103, "step": 77510 }, { "grad_norm": 0.07388008385896683, "learning_rate": 1.3192534300948811e-05, "loss": 0.0103, "step": 77520 }, { "grad_norm": 0.2708650231361389, "learning_rate": 1.3181345313077675e-05, "loss": 0.0153, "step": 77530 }, { "grad_norm": 0.2984425723552704, "learning_rate": 1.3170160351639243e-05, "loss": 0.0143, "step": 77540 }, { "grad_norm": 0.16462886333465576, "learning_rate": 1.3158979417856721e-05, "loss": 0.0171, "step": 77550 }, { "grad_norm": 0.16501833498477936, "learning_rate": 1.31478025129528e-05, "loss": 0.0119, "step": 77560 }, { "grad_norm": 0.3522280752658844, "learning_rate": 1.3136629638149794e-05, "loss": 0.0143, "step": 77570 }, { "grad_norm": 0.24922263622283936, "learning_rate": 1.3125460794669547e-05, "loss": 0.0174, "step": 77580 }, { "grad_norm": 0.10511357337236404, "learning_rate": 1.3114295983733483e-05, "loss": 0.017, "step": 77590 }, { "grad_norm": 0.21064385771751404, "learning_rate": 1.3103135206562534e-05, "loss": 0.0108, "step": 77600 }, { "grad_norm": 0.04794052243232727, "learning_rate": 1.3091978464377264e-05, "loss": 0.0227, "step": 77610 }, { "grad_norm": 0.4197816848754883, "learning_rate": 1.3080825758397724e-05, "loss": 0.0127, "step": 77620 }, { "grad_norm": 0.42494502663612366, "learning_rate": 1.3069677089843568e-05, "loss": 0.0201, "step": 77630 }, { "grad_norm": 0.07952336221933365, "learning_rate": 1.305853245993401e-05, "loss": 0.0156, "step": 77640 }, { "grad_norm": 0.09547832608222961, "learning_rate": 1.3047391869887809e-05, "loss": 0.0124, "step": 77650 }, { "grad_norm": 0.10838740319013596, "learning_rate": 1.3036255320923263e-05, "loss": 0.0156, "step": 77660 }, { "grad_norm": 0.3384505808353424, "learning_rate": 1.3025122814258273e-05, "loss": 0.0136, "step": 77670 }, { "grad_norm": 0.20753271877765656, "learning_rate": 1.3013994351110249e-05, "loss": 0.0146, "step": 77680 }, { "grad_norm": 0.08141837269067764, "learning_rate": 1.3002869932696194e-05, "loss": 0.0093, "step": 77690 }, { "grad_norm": 0.09319493919610977, "learning_rate": 1.2991749560232664e-05, "loss": 0.0099, "step": 77700 }, { "grad_norm": 0.1986423134803772, "learning_rate": 1.2980633234935768e-05, "loss": 0.0189, "step": 77710 }, { "grad_norm": 0.12280551344156265, "learning_rate": 1.2969520958021153e-05, "loss": 0.0129, "step": 77720 }, { "grad_norm": 0.0751294195652008, "learning_rate": 1.295841273070405e-05, "loss": 0.0082, "step": 77730 }, { "grad_norm": 0.13789041340351105, "learning_rate": 1.2947308554199255e-05, "loss": 0.0076, "step": 77740 }, { "grad_norm": 0.11238544434309006, "learning_rate": 1.2936208429721075e-05, "loss": 0.016, "step": 77750 }, { "grad_norm": 0.10377560555934906, "learning_rate": 1.2925112358483416e-05, "loss": 0.0144, "step": 77760 }, { "grad_norm": 0.10182669013738632, "learning_rate": 1.2914020341699728e-05, "loss": 0.0151, "step": 77770 }, { "grad_norm": 0.18511247634887695, "learning_rate": 1.2902932380583032e-05, "loss": 0.0146, "step": 77780 }, { "grad_norm": 0.18504014611244202, "learning_rate": 1.2891848476345863e-05, "loss": 0.01, "step": 77790 }, { "grad_norm": 0.2649224102497101, "learning_rate": 1.2880768630200368e-05, "loss": 0.0154, "step": 77800 }, { "grad_norm": 0.41526859998703003, "learning_rate": 1.286969284335819e-05, "loss": 0.0128, "step": 77810 }, { "grad_norm": 0.10377220809459686, "learning_rate": 1.2858621117030578e-05, "loss": 0.0091, "step": 77820 }, { "grad_norm": 0.2966633439064026, "learning_rate": 1.2847553452428324e-05, "loss": 0.0092, "step": 77830 }, { "grad_norm": 0.1671009659767151, "learning_rate": 1.283648985076178e-05, "loss": 0.0083, "step": 77840 }, { "grad_norm": 0.09563370794057846, "learning_rate": 1.2825430313240817e-05, "loss": 0.0106, "step": 77850 }, { "grad_norm": 0.21203452348709106, "learning_rate": 1.281437484107491e-05, "loss": 0.0194, "step": 77860 }, { "grad_norm": 0.11744243651628494, "learning_rate": 1.2803323435473058e-05, "loss": 0.0214, "step": 77870 }, { "grad_norm": 0.29851076006889343, "learning_rate": 1.2792276097643841e-05, "loss": 0.0159, "step": 77880 }, { "grad_norm": 0.1625443398952484, "learning_rate": 1.2781232828795342e-05, "loss": 0.0179, "step": 77890 }, { "grad_norm": 0.3088884651660919, "learning_rate": 1.2770193630135292e-05, "loss": 0.0169, "step": 77900 }, { "grad_norm": 0.25580477714538574, "learning_rate": 1.2759158502870877e-05, "loss": 0.0169, "step": 77910 }, { "grad_norm": 0.24507831037044525, "learning_rate": 1.2748127448208913e-05, "loss": 0.0134, "step": 77920 }, { "grad_norm": 0.4180906116962433, "learning_rate": 1.2737100467355706e-05, "loss": 0.0224, "step": 77930 }, { "grad_norm": 0.17449945211410522, "learning_rate": 1.2726077561517169e-05, "loss": 0.0211, "step": 77940 }, { "grad_norm": 0.15502393245697021, "learning_rate": 1.2715058731898755e-05, "loss": 0.0133, "step": 77950 }, { "grad_norm": 0.3140925168991089, "learning_rate": 1.2704043979705471e-05, "loss": 0.0132, "step": 77960 }, { "grad_norm": 0.34218552708625793, "learning_rate": 1.269303330614185e-05, "loss": 0.0093, "step": 77970 }, { "grad_norm": 0.03969372808933258, "learning_rate": 1.2682026712412015e-05, "loss": 0.0081, "step": 77980 }, { "grad_norm": 0.1824483424425125, "learning_rate": 1.2671024199719644e-05, "loss": 0.0072, "step": 77990 }, { "grad_norm": 0.043135471642017365, "learning_rate": 1.2660025769267936e-05, "loss": 0.0103, "step": 78000 }, { "grad_norm": 0.2593247592449188, "learning_rate": 1.2649031422259667e-05, "loss": 0.0189, "step": 78010 }, { "grad_norm": 0.14606092870235443, "learning_rate": 1.2638041159897163e-05, "loss": 0.0078, "step": 78020 }, { "grad_norm": 0.18347512185573578, "learning_rate": 1.262705498338232e-05, "loss": 0.0109, "step": 78030 }, { "grad_norm": 0.23579683899879456, "learning_rate": 1.2616072893916542e-05, "loss": 0.0187, "step": 78040 }, { "grad_norm": 0.16793501377105713, "learning_rate": 1.2605094892700836e-05, "loss": 0.015, "step": 78050 }, { "grad_norm": 0.09650961309671402, "learning_rate": 1.259412098093572e-05, "loss": 0.0095, "step": 78060 }, { "grad_norm": 0.09859127551317215, "learning_rate": 1.258315115982131e-05, "loss": 0.0082, "step": 78070 }, { "grad_norm": 0.11856595426797867, "learning_rate": 1.2572185430557203e-05, "loss": 0.007, "step": 78080 }, { "grad_norm": 0.1116795465350151, "learning_rate": 1.2561223794342659e-05, "loss": 0.0096, "step": 78090 }, { "grad_norm": 0.16588276624679565, "learning_rate": 1.255026625237637e-05, "loss": 0.0138, "step": 78100 }, { "grad_norm": 1.6196608543395996, "learning_rate": 1.2539312805856678e-05, "loss": 0.0151, "step": 78110 }, { "grad_norm": 0.14370490610599518, "learning_rate": 1.25283634559814e-05, "loss": 0.0094, "step": 78120 }, { "grad_norm": 0.10902254283428192, "learning_rate": 1.2517418203947972e-05, "loss": 0.0127, "step": 78130 }, { "grad_norm": 0.08001898974180222, "learning_rate": 1.2506477050953308e-05, "loss": 0.0094, "step": 78140 }, { "grad_norm": 0.27995556592941284, "learning_rate": 1.2495539998193972e-05, "loss": 0.0131, "step": 78150 }, { "grad_norm": 0.11900624632835388, "learning_rate": 1.2484607046865975e-05, "loss": 0.0103, "step": 78160 }, { "grad_norm": 0.2436979115009308, "learning_rate": 1.2473678198164968e-05, "loss": 0.0127, "step": 78170 }, { "grad_norm": 0.2061927765607834, "learning_rate": 1.2462753453286075e-05, "loss": 0.0142, "step": 78180 }, { "grad_norm": 0.10828092694282532, "learning_rate": 1.2451832813424031e-05, "loss": 0.0237, "step": 78190 }, { "grad_norm": 0.24160340428352356, "learning_rate": 1.2440916279773095e-05, "loss": 0.0144, "step": 78200 }, { "grad_norm": 0.18841689825057983, "learning_rate": 1.2430003853527101e-05, "loss": 0.0098, "step": 78210 }, { "grad_norm": 0.3176780045032501, "learning_rate": 1.2419095535879383e-05, "loss": 0.0176, "step": 78220 }, { "grad_norm": 0.21209031343460083, "learning_rate": 1.2408191328022878e-05, "loss": 0.0174, "step": 78230 }, { "grad_norm": 0.21728329360485077, "learning_rate": 1.2397291231150066e-05, "loss": 0.0119, "step": 78240 }, { "grad_norm": 0.18670058250427246, "learning_rate": 1.2386395246452937e-05, "loss": 0.013, "step": 78250 }, { "grad_norm": 0.26700520515441895, "learning_rate": 1.2375503375123083e-05, "loss": 0.0122, "step": 78260 }, { "grad_norm": 0.1280948966741562, "learning_rate": 1.2364615618351583e-05, "loss": 0.0164, "step": 78270 }, { "grad_norm": 0.5760535001754761, "learning_rate": 1.2353731977329169e-05, "loss": 0.017, "step": 78280 }, { "grad_norm": 0.10421310365200043, "learning_rate": 1.2342852453246007e-05, "loss": 0.0114, "step": 78290 }, { "grad_norm": 0.3603266775608063, "learning_rate": 1.2331977047291898e-05, "loss": 0.0229, "step": 78300 }, { "grad_norm": 0.1576090157032013, "learning_rate": 1.2321105760656137e-05, "loss": 0.0143, "step": 78310 }, { "grad_norm": 0.04522433131933212, "learning_rate": 1.2310238594527613e-05, "loss": 0.0129, "step": 78320 }, { "grad_norm": 0.2803988754749298, "learning_rate": 1.2299375550094693e-05, "loss": 0.0151, "step": 78330 }, { "grad_norm": 0.09359557926654816, "learning_rate": 1.2288516628545421e-05, "loss": 0.0093, "step": 78340 }, { "grad_norm": 0.13480855524539948, "learning_rate": 1.2277661831067255e-05, "loss": 0.0134, "step": 78350 }, { "grad_norm": 0.14077837765216827, "learning_rate": 1.2266811158847286e-05, "loss": 0.0167, "step": 78360 }, { "grad_norm": 0.16286775469779968, "learning_rate": 1.2255964613072107e-05, "loss": 0.0105, "step": 78370 }, { "grad_norm": 0.11000676453113556, "learning_rate": 1.2245122194927905e-05, "loss": 0.0089, "step": 78380 }, { "grad_norm": 0.1203024834394455, "learning_rate": 1.2234283905600364e-05, "loss": 0.0115, "step": 78390 }, { "grad_norm": 0.1843315213918686, "learning_rate": 1.222344974627475e-05, "loss": 0.016, "step": 78400 }, { "grad_norm": 0.14763301610946655, "learning_rate": 1.2212619718135875e-05, "loss": 0.0092, "step": 78410 }, { "grad_norm": 0.18304993212223053, "learning_rate": 1.2201793822368118e-05, "loss": 0.0054, "step": 78420 }, { "grad_norm": 0.08699808269739151, "learning_rate": 1.2190972060155337e-05, "loss": 0.0085, "step": 78430 }, { "grad_norm": 0.24724814295768738, "learning_rate": 1.2180154432681013e-05, "loss": 0.0108, "step": 78440 }, { "grad_norm": 0.1408897191286087, "learning_rate": 1.2169340941128148e-05, "loss": 0.0103, "step": 78450 }, { "grad_norm": 0.17650575935840607, "learning_rate": 1.2158531586679267e-05, "loss": 0.015, "step": 78460 }, { "grad_norm": 0.40639740228652954, "learning_rate": 1.2147726370516477e-05, "loss": 0.0134, "step": 78470 }, { "grad_norm": 0.1603374034166336, "learning_rate": 1.2136925293821422e-05, "loss": 0.0119, "step": 78480 }, { "grad_norm": 0.2021041363477707, "learning_rate": 1.2126128357775306e-05, "loss": 0.0101, "step": 78490 }, { "grad_norm": 0.13692717254161835, "learning_rate": 1.2115335563558838e-05, "loss": 0.0119, "step": 78500 }, { "grad_norm": 0.38393738865852356, "learning_rate": 1.2104546912352327e-05, "loss": 0.0114, "step": 78510 }, { "grad_norm": 0.4379000961780548, "learning_rate": 1.2093762405335557e-05, "loss": 0.0098, "step": 78520 }, { "grad_norm": 0.09763957560062408, "learning_rate": 1.2082982043687974e-05, "loss": 0.0118, "step": 78530 }, { "grad_norm": 0.06895449012517929, "learning_rate": 1.207220582858845e-05, "loss": 0.0094, "step": 78540 }, { "grad_norm": 0.2092028558254242, "learning_rate": 1.206143376121549e-05, "loss": 0.0092, "step": 78550 }, { "grad_norm": 1.0683668851852417, "learning_rate": 1.2050665842747078e-05, "loss": 0.0067, "step": 78560 }, { "grad_norm": 0.1382080614566803, "learning_rate": 1.2039902074360804e-05, "loss": 0.0158, "step": 78570 }, { "grad_norm": 0.1450527459383011, "learning_rate": 1.2029142457233755e-05, "loss": 0.0094, "step": 78580 }, { "grad_norm": 0.10551133751869202, "learning_rate": 1.2018386992542601e-05, "loss": 0.0077, "step": 78590 }, { "grad_norm": 0.1121029257774353, "learning_rate": 1.2007635681463542e-05, "loss": 0.0105, "step": 78600 }, { "grad_norm": 0.3243042230606079, "learning_rate": 1.1996888525172334e-05, "loss": 0.0136, "step": 78610 }, { "grad_norm": 0.08864568918943405, "learning_rate": 1.198614552484425e-05, "loss": 0.0075, "step": 78620 }, { "grad_norm": 0.2564176917076111, "learning_rate": 1.197540668165415e-05, "loss": 0.0177, "step": 78630 }, { "grad_norm": 0.12023302912712097, "learning_rate": 1.1964671996776395e-05, "loss": 0.0103, "step": 78640 }, { "grad_norm": 0.13753046095371246, "learning_rate": 1.1953941471384922e-05, "loss": 0.0123, "step": 78650 }, { "grad_norm": 0.26593855023384094, "learning_rate": 1.1943215106653205e-05, "loss": 0.0203, "step": 78660 }, { "grad_norm": 0.17208491265773773, "learning_rate": 1.193249290375429e-05, "loss": 0.0109, "step": 78670 }, { "grad_norm": 0.08259040862321854, "learning_rate": 1.1921774863860696e-05, "loss": 0.0159, "step": 78680 }, { "grad_norm": 0.10909519344568253, "learning_rate": 1.1911060988144556e-05, "loss": 0.0096, "step": 78690 }, { "grad_norm": 0.27808451652526855, "learning_rate": 1.1900351277777533e-05, "loss": 0.0149, "step": 78700 }, { "grad_norm": 0.16143907606601715, "learning_rate": 1.1889645733930798e-05, "loss": 0.0112, "step": 78710 }, { "grad_norm": 0.09595349431037903, "learning_rate": 1.1878944357775107e-05, "loss": 0.0085, "step": 78720 }, { "grad_norm": 0.11784037947654724, "learning_rate": 1.1868247150480749e-05, "loss": 0.0092, "step": 78730 }, { "grad_norm": 0.0998603031039238, "learning_rate": 1.1857554113217567e-05, "loss": 0.0082, "step": 78740 }, { "grad_norm": 0.16498196125030518, "learning_rate": 1.1846865247154903e-05, "loss": 0.0128, "step": 78750 }, { "grad_norm": 0.053126879036426544, "learning_rate": 1.1836180553461711e-05, "loss": 0.015, "step": 78760 }, { "grad_norm": 0.07679043710231781, "learning_rate": 1.1825500033306419e-05, "loss": 0.0269, "step": 78770 }, { "grad_norm": 0.21632227301597595, "learning_rate": 1.1814823687857052e-05, "loss": 0.0104, "step": 78780 }, { "grad_norm": 0.1782638430595398, "learning_rate": 1.1804151518281153e-05, "loss": 0.0089, "step": 78790 }, { "grad_norm": 0.10476365685462952, "learning_rate": 1.179348352574583e-05, "loss": 0.0293, "step": 78800 }, { "grad_norm": 0.12840458750724792, "learning_rate": 1.1782819711417697e-05, "loss": 0.0143, "step": 78810 }, { "grad_norm": 0.12318983674049377, "learning_rate": 1.1772160076462951e-05, "loss": 0.0105, "step": 78820 }, { "grad_norm": 0.2910041809082031, "learning_rate": 1.1761504622047293e-05, "loss": 0.014, "step": 78830 }, { "grad_norm": 0.15124031901359558, "learning_rate": 1.1750853349335999e-05, "loss": 0.0094, "step": 78840 }, { "grad_norm": 0.15876562893390656, "learning_rate": 1.1740206259493874e-05, "loss": 0.0123, "step": 78850 }, { "grad_norm": 0.08887223154306412, "learning_rate": 1.1729563353685286e-05, "loss": 0.0117, "step": 78860 }, { "grad_norm": 0.9269435405731201, "learning_rate": 1.1718924633074091e-05, "loss": 0.008, "step": 78870 }, { "grad_norm": 0.2046695351600647, "learning_rate": 1.1708290098823765e-05, "loss": 0.0105, "step": 78880 }, { "grad_norm": 0.2646557092666626, "learning_rate": 1.1697659752097245e-05, "loss": 0.0094, "step": 78890 }, { "grad_norm": 0.13886745274066925, "learning_rate": 1.1687033594057068e-05, "loss": 0.0249, "step": 78900 }, { "grad_norm": 0.22305303812026978, "learning_rate": 1.1676411625865297e-05, "loss": 0.0071, "step": 78910 }, { "grad_norm": 0.25011277198791504, "learning_rate": 1.1665793848683542e-05, "loss": 0.0115, "step": 78920 }, { "grad_norm": 0.1316991150379181, "learning_rate": 1.1655180263672927e-05, "loss": 0.0262, "step": 78930 }, { "grad_norm": 0.3393557667732239, "learning_rate": 1.1644570871994142e-05, "loss": 0.0142, "step": 78940 }, { "grad_norm": 0.10703899711370468, "learning_rate": 1.163396567480744e-05, "loss": 0.0117, "step": 78950 }, { "grad_norm": 0.13948631286621094, "learning_rate": 1.1623364673272552e-05, "loss": 0.0121, "step": 78960 }, { "grad_norm": 0.12223530560731888, "learning_rate": 1.1612767868548807e-05, "loss": 0.0116, "step": 78970 }, { "grad_norm": 0.364541620016098, "learning_rate": 1.1602175261795057e-05, "loss": 0.0156, "step": 78980 }, { "grad_norm": 0.1310531049966812, "learning_rate": 1.1591586854169696e-05, "loss": 0.0127, "step": 78990 }, { "grad_norm": 0.32037314772605896, "learning_rate": 1.158100264683064e-05, "loss": 0.0126, "step": 79000 }, { "grad_norm": 0.09653201699256897, "learning_rate": 1.1570422640935386e-05, "loss": 0.011, "step": 79010 }, { "grad_norm": 0.2574973702430725, "learning_rate": 1.155984683764092e-05, "loss": 0.0147, "step": 79020 }, { "grad_norm": 0.1535269021987915, "learning_rate": 1.1549275238103812e-05, "loss": 0.0083, "step": 79030 }, { "grad_norm": 0.37260669469833374, "learning_rate": 1.1538707843480152e-05, "loss": 0.0183, "step": 79040 }, { "grad_norm": 0.2092725932598114, "learning_rate": 1.1528144654925593e-05, "loss": 0.0133, "step": 79050 }, { "grad_norm": 0.14740343391895294, "learning_rate": 1.1517585673595277e-05, "loss": 0.015, "step": 79060 }, { "grad_norm": 0.10839452594518661, "learning_rate": 1.150703090064395e-05, "loss": 0.0127, "step": 79070 }, { "grad_norm": 0.09357787668704987, "learning_rate": 1.1496480337225835e-05, "loss": 0.015, "step": 79080 }, { "grad_norm": 0.11835174262523651, "learning_rate": 1.1485933984494751e-05, "loss": 0.011, "step": 79090 }, { "grad_norm": 0.15991246700286865, "learning_rate": 1.1475391843604e-05, "loss": 0.0107, "step": 79100 }, { "grad_norm": 0.18157561123371124, "learning_rate": 1.1464853915706503e-05, "loss": 0.0229, "step": 79110 }, { "grad_norm": 0.24213998019695282, "learning_rate": 1.1454320201954626e-05, "loss": 0.0103, "step": 79120 }, { "grad_norm": 0.5039203763008118, "learning_rate": 1.1443790703500356e-05, "loss": 0.0086, "step": 79130 }, { "grad_norm": 0.08750438690185547, "learning_rate": 1.1433265421495154e-05, "loss": 0.0115, "step": 79140 }, { "grad_norm": 0.2379736751317978, "learning_rate": 1.1422744357090065e-05, "loss": 0.0261, "step": 79150 }, { "grad_norm": 0.1767793893814087, "learning_rate": 1.1412227511435647e-05, "loss": 0.0121, "step": 79160 }, { "grad_norm": 0.37379732728004456, "learning_rate": 1.1401714885682025e-05, "loss": 0.0139, "step": 79170 }, { "grad_norm": 0.12325582653284073, "learning_rate": 1.1391206480978823e-05, "loss": 0.0124, "step": 79180 }, { "grad_norm": 0.04981432855129242, "learning_rate": 1.1380702298475227e-05, "loss": 0.0136, "step": 79190 }, { "grad_norm": 0.07416058331727982, "learning_rate": 1.1370202339319985e-05, "loss": 0.0099, "step": 79200 }, { "grad_norm": 0.17389923334121704, "learning_rate": 1.1359706604661319e-05, "loss": 0.0184, "step": 79210 }, { "grad_norm": 0.41600438952445984, "learning_rate": 1.1349215095647042e-05, "loss": 0.0132, "step": 79220 }, { "grad_norm": 0.10227824747562408, "learning_rate": 1.1338727813424493e-05, "loss": 0.0099, "step": 79230 }, { "grad_norm": 0.2580260634422302, "learning_rate": 1.132824475914056e-05, "loss": 0.0126, "step": 79240 }, { "grad_norm": 0.43967941403388977, "learning_rate": 1.131776593394162e-05, "loss": 0.0126, "step": 79250 }, { "grad_norm": 0.2853566110134125, "learning_rate": 1.1307291338973657e-05, "loss": 0.0109, "step": 79260 }, { "grad_norm": 0.1440933644771576, "learning_rate": 1.1296820975382121e-05, "loss": 0.0121, "step": 79270 }, { "grad_norm": 0.08338481932878494, "learning_rate": 1.1286354844312074e-05, "loss": 0.0193, "step": 79280 }, { "grad_norm": 0.05665387213230133, "learning_rate": 1.1275892946908023e-05, "loss": 0.0118, "step": 79290 }, { "grad_norm": 0.2600186765193939, "learning_rate": 1.126543528431413e-05, "loss": 0.0112, "step": 79300 }, { "grad_norm": 0.34194204211235046, "learning_rate": 1.125498185767398e-05, "loss": 0.011, "step": 79310 }, { "grad_norm": 0.11027223616838455, "learning_rate": 1.1244532668130781e-05, "loss": 0.0117, "step": 79320 }, { "grad_norm": 0.16030102968215942, "learning_rate": 1.1234087716827203e-05, "loss": 0.0179, "step": 79330 }, { "grad_norm": 0.31907230615615845, "learning_rate": 1.1223647004905524e-05, "loss": 0.0149, "step": 79340 }, { "grad_norm": 0.16772033274173737, "learning_rate": 1.1213210533507485e-05, "loss": 0.0174, "step": 79350 }, { "grad_norm": 0.3438000977039337, "learning_rate": 1.1202778303774447e-05, "loss": 0.0201, "step": 79360 }, { "grad_norm": 0.2272367924451828, "learning_rate": 1.1192350316847234e-05, "loss": 0.0147, "step": 79370 }, { "grad_norm": 0.3295592963695526, "learning_rate": 1.1181926573866258e-05, "loss": 0.0166, "step": 79380 }, { "grad_norm": 0.3397584557533264, "learning_rate": 1.1171507075971416e-05, "loss": 0.0174, "step": 79390 }, { "grad_norm": 0.2575532793998718, "learning_rate": 1.1161091824302177e-05, "loss": 0.0114, "step": 79400 }, { "grad_norm": 0.3997340202331543, "learning_rate": 1.115068081999755e-05, "loss": 0.0173, "step": 79410 }, { "grad_norm": 0.24721789360046387, "learning_rate": 1.1140274064196071e-05, "loss": 0.0122, "step": 79420 }, { "grad_norm": 0.11219660937786102, "learning_rate": 1.112987155803578e-05, "loss": 0.0154, "step": 79430 }, { "grad_norm": 0.1618821769952774, "learning_rate": 1.1119473302654298e-05, "loss": 0.0178, "step": 79440 }, { "grad_norm": 0.11532974243164062, "learning_rate": 1.1109079299188769e-05, "loss": 0.0126, "step": 79450 }, { "grad_norm": 0.319480836391449, "learning_rate": 1.1098689548775847e-05, "loss": 0.0148, "step": 79460 }, { "grad_norm": 0.5782322287559509, "learning_rate": 1.1088304052551762e-05, "loss": 0.0098, "step": 79470 }, { "grad_norm": 0.345060259103775, "learning_rate": 1.1077922811652215e-05, "loss": 0.0097, "step": 79480 }, { "grad_norm": 0.07024500519037247, "learning_rate": 1.1067545827212534e-05, "loss": 0.011, "step": 79490 }, { "grad_norm": 0.4170837700366974, "learning_rate": 1.1057173100367496e-05, "loss": 0.01, "step": 79500 }, { "grad_norm": 0.08654584735631943, "learning_rate": 1.1046804632251474e-05, "loss": 0.0152, "step": 79510 }, { "grad_norm": 0.1745867133140564, "learning_rate": 1.1036440423998318e-05, "loss": 0.0098, "step": 79520 }, { "grad_norm": 0.2688582241535187, "learning_rate": 1.1026080476741469e-05, "loss": 0.0161, "step": 79530 }, { "grad_norm": 0.15974363684654236, "learning_rate": 1.1015724791613836e-05, "loss": 0.0129, "step": 79540 }, { "grad_norm": 0.2397618442773819, "learning_rate": 1.1005373369747951e-05, "loss": 0.0094, "step": 79550 }, { "grad_norm": 0.14662805199623108, "learning_rate": 1.09950262122758e-05, "loss": 0.0118, "step": 79560 }, { "grad_norm": 0.047070953994989395, "learning_rate": 1.0984683320328948e-05, "loss": 0.0101, "step": 79570 }, { "grad_norm": 0.1403159499168396, "learning_rate": 1.0974344695038458e-05, "loss": 0.0106, "step": 79580 }, { "grad_norm": 0.4605048894882202, "learning_rate": 1.0964010337534968e-05, "loss": 0.0179, "step": 79590 }, { "grad_norm": 0.14235767722129822, "learning_rate": 1.0953680248948611e-05, "loss": 0.0099, "step": 79600 }, { "grad_norm": 0.1321968138217926, "learning_rate": 1.0943354430409069e-05, "loss": 0.0081, "step": 79610 }, { "grad_norm": 0.33182016015052795, "learning_rate": 1.0933032883045574e-05, "loss": 0.0135, "step": 79620 }, { "grad_norm": 0.1513752043247223, "learning_rate": 1.0922715607986872e-05, "loss": 0.0118, "step": 79630 }, { "grad_norm": 0.19182078540325165, "learning_rate": 1.0912402606361234e-05, "loss": 0.0068, "step": 79640 }, { "grad_norm": 0.12839758396148682, "learning_rate": 1.090209387929647e-05, "loss": 0.0112, "step": 79650 }, { "grad_norm": 0.26125046610832214, "learning_rate": 1.0891789427919957e-05, "loss": 0.0165, "step": 79660 }, { "grad_norm": 0.13473421335220337, "learning_rate": 1.088148925335854e-05, "loss": 0.0161, "step": 79670 }, { "grad_norm": 0.21066837012767792, "learning_rate": 1.087119335673864e-05, "loss": 0.0167, "step": 79680 }, { "grad_norm": 0.1864287108182907, "learning_rate": 1.0860901739186208e-05, "loss": 0.0094, "step": 79690 }, { "grad_norm": 0.11061590909957886, "learning_rate": 1.085061440182673e-05, "loss": 0.0154, "step": 79700 }, { "grad_norm": 0.18885421752929688, "learning_rate": 1.0840331345785193e-05, "loss": 0.009, "step": 79710 }, { "grad_norm": 0.12735389173030853, "learning_rate": 1.083005257218615e-05, "loss": 0.0079, "step": 79720 }, { "grad_norm": 0.1758720725774765, "learning_rate": 1.0819778082153642e-05, "loss": 0.0122, "step": 79730 }, { "grad_norm": 0.18970629572868347, "learning_rate": 1.0809507876811325e-05, "loss": 0.0095, "step": 79740 }, { "grad_norm": 0.07745982706546783, "learning_rate": 1.0799241957282292e-05, "loss": 0.0055, "step": 79750 }, { "grad_norm": 0.13587379455566406, "learning_rate": 1.078898032468923e-05, "loss": 0.0088, "step": 79760 }, { "grad_norm": 0.11479125916957855, "learning_rate": 1.0778722980154315e-05, "loss": 0.015, "step": 79770 }, { "grad_norm": 0.19341568648815155, "learning_rate": 1.0768469924799302e-05, "loss": 0.0137, "step": 79780 }, { "grad_norm": 0.10024911910295486, "learning_rate": 1.0758221159745419e-05, "loss": 0.013, "step": 79790 }, { "grad_norm": 0.09372775256633759, "learning_rate": 1.074797668611347e-05, "loss": 0.0132, "step": 79800 }, { "grad_norm": 0.1846364587545395, "learning_rate": 1.0737736505023777e-05, "loss": 0.0138, "step": 79810 }, { "grad_norm": 0.29667285084724426, "learning_rate": 1.0727500617596203e-05, "loss": 0.0158, "step": 79820 }, { "grad_norm": 0.09944027662277222, "learning_rate": 1.0717269024950104e-05, "loss": 0.0168, "step": 79830 }, { "grad_norm": 0.05820821225643158, "learning_rate": 1.070704172820442e-05, "loss": 0.0109, "step": 79840 }, { "grad_norm": 0.187273308634758, "learning_rate": 1.0696818728477559e-05, "loss": 0.0148, "step": 79850 }, { "grad_norm": 0.18653592467308044, "learning_rate": 1.0686600026887511e-05, "loss": 0.0107, "step": 79860 }, { "grad_norm": 0.12062542140483856, "learning_rate": 1.0676385624551777e-05, "loss": 0.0145, "step": 79870 }, { "grad_norm": 0.15560315549373627, "learning_rate": 1.0666175522587402e-05, "loss": 0.008, "step": 79880 }, { "grad_norm": 0.16660773754119873, "learning_rate": 1.0655969722110925e-05, "loss": 0.0088, "step": 79890 }, { "grad_norm": 0.11404328793287277, "learning_rate": 1.0645768224238439e-05, "loss": 0.0126, "step": 79900 }, { "grad_norm": 0.36209264397621155, "learning_rate": 1.063557103008559e-05, "loss": 0.0126, "step": 79910 }, { "grad_norm": 0.37193062901496887, "learning_rate": 1.0625378140767494e-05, "loss": 0.0126, "step": 79920 }, { "grad_norm": 0.39527663588523865, "learning_rate": 1.0615189557398846e-05, "loss": 0.0141, "step": 79930 }, { "grad_norm": 0.14818648993968964, "learning_rate": 1.060500528109386e-05, "loss": 0.0186, "step": 79940 }, { "grad_norm": 0.17155244946479797, "learning_rate": 1.0594825312966272e-05, "loss": 0.0116, "step": 79950 }, { "grad_norm": 0.11553144454956055, "learning_rate": 1.0584649654129336e-05, "loss": 0.0197, "step": 79960 }, { "grad_norm": 0.13595668971538544, "learning_rate": 1.0574478305695863e-05, "loss": 0.0086, "step": 79970 }, { "grad_norm": 0.09894907474517822, "learning_rate": 1.0564311268778154e-05, "loss": 0.0114, "step": 79980 }, { "grad_norm": 0.385755330324173, "learning_rate": 1.0554148544488073e-05, "loss": 0.0144, "step": 79990 }, { "grad_norm": 0.13264593482017517, "learning_rate": 1.0543990133937004e-05, "loss": 0.0109, "step": 80000 }, { "grad_norm": 0.12479732930660248, "learning_rate": 1.0533836038235868e-05, "loss": 0.006, "step": 80010 }, { "grad_norm": 0.030761683359742165, "learning_rate": 1.052368625849507e-05, "loss": 0.0123, "step": 80020 }, { "grad_norm": 0.30245617032051086, "learning_rate": 1.05135407958246e-05, "loss": 0.0081, "step": 80030 }, { "grad_norm": 0.3384578227996826, "learning_rate": 1.0503399651333934e-05, "loss": 0.0148, "step": 80040 }, { "grad_norm": 0.25103873014450073, "learning_rate": 1.04932628261321e-05, "loss": 0.014, "step": 80050 }, { "grad_norm": 0.12722405791282654, "learning_rate": 1.0483130321327644e-05, "loss": 0.011, "step": 80060 }, { "grad_norm": 0.1713186800479889, "learning_rate": 1.0473002138028654e-05, "loss": 0.0124, "step": 80070 }, { "grad_norm": 0.18958351016044617, "learning_rate": 1.0462878277342713e-05, "loss": 0.0097, "step": 80080 }, { "grad_norm": 0.0788152664899826, "learning_rate": 1.0452758740376972e-05, "loss": 0.0108, "step": 80090 }, { "grad_norm": 0.08723796159029007, "learning_rate": 1.0442643528238067e-05, "loss": 0.0068, "step": 80100 }, { "grad_norm": 0.22351552546024323, "learning_rate": 1.0432532642032188e-05, "loss": 0.0115, "step": 80110 }, { "grad_norm": 0.3409793972969055, "learning_rate": 1.042242608286505e-05, "loss": 0.0102, "step": 80120 }, { "grad_norm": 0.21879741549491882, "learning_rate": 1.041232385184191e-05, "loss": 0.0086, "step": 80130 }, { "grad_norm": 0.3042048513889313, "learning_rate": 1.0402225950067495e-05, "loss": 0.011, "step": 80140 }, { "grad_norm": 0.17530925571918488, "learning_rate": 1.0392132378646119e-05, "loss": 0.0109, "step": 80150 }, { "grad_norm": 0.1506527066230774, "learning_rate": 1.0382043138681607e-05, "loss": 0.0082, "step": 80160 }, { "grad_norm": 0.23438985645771027, "learning_rate": 1.0371958231277284e-05, "loss": 0.0117, "step": 80170 }, { "grad_norm": 0.222019225358963, "learning_rate": 1.0361877657536024e-05, "loss": 0.009, "step": 80180 }, { "grad_norm": 0.19127091765403748, "learning_rate": 1.035180141856023e-05, "loss": 0.011, "step": 80190 }, { "grad_norm": 0.30664247274398804, "learning_rate": 1.0341729515451836e-05, "loss": 0.0137, "step": 80200 }, { "grad_norm": 0.17193327844142914, "learning_rate": 1.0331661949312265e-05, "loss": 0.01, "step": 80210 }, { "grad_norm": 0.22487159073352814, "learning_rate": 1.0321598721242514e-05, "loss": 0.0136, "step": 80220 }, { "grad_norm": 0.1987103521823883, "learning_rate": 1.0311539832343064e-05, "loss": 0.0116, "step": 80230 }, { "grad_norm": 0.1459064483642578, "learning_rate": 1.0301485283713942e-05, "loss": 0.0085, "step": 80240 }, { "grad_norm": 0.10661246627569199, "learning_rate": 1.0291435076454703e-05, "loss": 0.0152, "step": 80250 }, { "grad_norm": 0.21503634750843048, "learning_rate": 1.028138921166444e-05, "loss": 0.0144, "step": 80260 }, { "grad_norm": 0.07456027716398239, "learning_rate": 1.0271347690441719e-05, "loss": 0.0077, "step": 80270 }, { "grad_norm": 0.19386287033557892, "learning_rate": 1.0261310513884698e-05, "loss": 0.0126, "step": 80280 }, { "grad_norm": 0.17498062551021576, "learning_rate": 1.0251277683090998e-05, "loss": 0.0076, "step": 80290 }, { "grad_norm": 0.6593270301818848, "learning_rate": 1.0241249199157826e-05, "loss": 0.012, "step": 80300 }, { "grad_norm": 0.22523978352546692, "learning_rate": 1.0231225063181837e-05, "loss": 0.0063, "step": 80310 }, { "grad_norm": 0.1560254544019699, "learning_rate": 1.0221205276259316e-05, "loss": 0.0098, "step": 80320 }, { "grad_norm": 0.0543527826666832, "learning_rate": 1.0211189839485958e-05, "loss": 0.0112, "step": 80330 }, { "grad_norm": 0.194846510887146, "learning_rate": 1.0201178753957075e-05, "loss": 0.0098, "step": 80340 }, { "grad_norm": 0.22146666049957275, "learning_rate": 1.0191172020767432e-05, "loss": 0.0139, "step": 80350 }, { "grad_norm": 0.08535107225179672, "learning_rate": 1.0181169641011362e-05, "loss": 0.0112, "step": 80360 }, { "grad_norm": 0.2016090303659439, "learning_rate": 1.0171171615782721e-05, "loss": 0.0129, "step": 80370 }, { "grad_norm": 0.07443570345640182, "learning_rate": 1.0161177946174866e-05, "loss": 0.0093, "step": 80380 }, { "grad_norm": 0.2630588710308075, "learning_rate": 1.0151188633280706e-05, "loss": 0.0148, "step": 80390 }, { "grad_norm": 0.29173362255096436, "learning_rate": 1.0141203678192635e-05, "loss": 0.0142, "step": 80400 }, { "grad_norm": 0.10740210860967636, "learning_rate": 1.0131223082002612e-05, "loss": 0.0145, "step": 80410 }, { "grad_norm": 0.2123890072107315, "learning_rate": 1.012124684580208e-05, "loss": 0.0102, "step": 80420 }, { "grad_norm": 0.09124209731817245, "learning_rate": 1.0111274970682039e-05, "loss": 0.012, "step": 80430 }, { "grad_norm": 0.08433623611927032, "learning_rate": 1.010130745773299e-05, "loss": 0.0071, "step": 80440 }, { "grad_norm": 0.08699119091033936, "learning_rate": 1.0091344308044986e-05, "loss": 0.0075, "step": 80450 }, { "grad_norm": 0.17872199416160583, "learning_rate": 1.0081385522707554e-05, "loss": 0.0114, "step": 80460 }, { "grad_norm": 0.06500011682510376, "learning_rate": 1.0071431102809803e-05, "loss": 0.0103, "step": 80470 }, { "grad_norm": 0.07475769519805908, "learning_rate": 1.0061481049440297e-05, "loss": 0.0074, "step": 80480 }, { "grad_norm": 0.2209104299545288, "learning_rate": 1.0051535363687187e-05, "loss": 0.0117, "step": 80490 }, { "grad_norm": 0.17227892577648163, "learning_rate": 1.0041594046638087e-05, "loss": 0.016, "step": 80500 }, { "grad_norm": 0.14115744829177856, "learning_rate": 1.0031657099380204e-05, "loss": 0.0081, "step": 80510 }, { "grad_norm": 0.279546320438385, "learning_rate": 1.0021724523000204e-05, "loss": 0.0132, "step": 80520 }, { "grad_norm": 0.150425523519516, "learning_rate": 1.001179631858431e-05, "loss": 0.008, "step": 80530 }, { "grad_norm": 0.1729152500629425, "learning_rate": 1.0001872487218245e-05, "loss": 0.0157, "step": 80540 }, { "grad_norm": 0.19961747527122498, "learning_rate": 9.991953029987272e-06, "loss": 0.0124, "step": 80550 }, { "grad_norm": 0.28184327483177185, "learning_rate": 9.982037947976147e-06, "loss": 0.0149, "step": 80560 }, { "grad_norm": 0.18380622565746307, "learning_rate": 9.972127242269208e-06, "loss": 0.0176, "step": 80570 }, { "grad_norm": 0.3803093731403351, "learning_rate": 9.962220913950238e-06, "loss": 0.0166, "step": 80580 }, { "grad_norm": 0.19400331377983093, "learning_rate": 9.952318964102591e-06, "loss": 0.0122, "step": 80590 }, { "grad_norm": 0.23562978208065033, "learning_rate": 9.942421393809148e-06, "loss": 0.0141, "step": 80600 }, { "grad_norm": 0.5427162051200867, "learning_rate": 9.932528204152264e-06, "loss": 0.0144, "step": 80610 }, { "grad_norm": 0.11053886264562607, "learning_rate": 9.922639396213856e-06, "loss": 0.0161, "step": 80620 }, { "grad_norm": 0.15158814191818237, "learning_rate": 9.912754971075344e-06, "loss": 0.0142, "step": 80630 }, { "grad_norm": 0.09436242282390594, "learning_rate": 9.902874929817696e-06, "loss": 0.0137, "step": 80640 }, { "grad_norm": 0.2731402814388275, "learning_rate": 9.892999273521342e-06, "loss": 0.0118, "step": 80650 }, { "grad_norm": 0.24364744126796722, "learning_rate": 9.883128003266302e-06, "loss": 0.0096, "step": 80660 }, { "grad_norm": 0.187132328748703, "learning_rate": 9.87326112013206e-06, "loss": 0.0081, "step": 80670 }, { "grad_norm": 0.261104553937912, "learning_rate": 9.863398625197662e-06, "loss": 0.0135, "step": 80680 }, { "grad_norm": 0.1657317876815796, "learning_rate": 9.853540519541615e-06, "loss": 0.0056, "step": 80690 }, { "grad_norm": 0.1824265718460083, "learning_rate": 9.843686804242053e-06, "loss": 0.0199, "step": 80700 }, { "grad_norm": 0.1563863605260849, "learning_rate": 9.833837480376506e-06, "loss": 0.0153, "step": 80710 }, { "grad_norm": 0.11367642879486084, "learning_rate": 9.82399254902212e-06, "loss": 0.0121, "step": 80720 }, { "grad_norm": 0.38930031657218933, "learning_rate": 9.814152011255494e-06, "loss": 0.0106, "step": 80730 }, { "grad_norm": 0.1541949063539505, "learning_rate": 9.804315868152797e-06, "loss": 0.0111, "step": 80740 }, { "grad_norm": 0.45390912890434265, "learning_rate": 9.79448412078966e-06, "loss": 0.0153, "step": 80750 }, { "grad_norm": 0.11716027557849884, "learning_rate": 9.784656770241312e-06, "loss": 0.0132, "step": 80760 }, { "grad_norm": 0.10120512545108795, "learning_rate": 9.77483381758243e-06, "loss": 0.0081, "step": 80770 }, { "grad_norm": 0.23957134783267975, "learning_rate": 9.76501526388725e-06, "loss": 0.0141, "step": 80780 }, { "grad_norm": 0.35528677701950073, "learning_rate": 9.7552011102295e-06, "loss": 0.0238, "step": 80790 }, { "grad_norm": 0.12149540334939957, "learning_rate": 9.745391357682449e-06, "loss": 0.0113, "step": 80800 }, { "grad_norm": 0.11700531095266342, "learning_rate": 9.735586007318875e-06, "loss": 0.0162, "step": 80810 }, { "grad_norm": 0.23555801808834076, "learning_rate": 9.725785060211096e-06, "loss": 0.0109, "step": 80820 }, { "grad_norm": 0.08862916380167007, "learning_rate": 9.715988517430896e-06, "loss": 0.012, "step": 80830 }, { "grad_norm": 0.17297707498073578, "learning_rate": 9.706196380049625e-06, "loss": 0.0107, "step": 80840 }, { "grad_norm": 0.38539230823516846, "learning_rate": 9.69640864913815e-06, "loss": 0.012, "step": 80850 }, { "grad_norm": 0.08975504338741302, "learning_rate": 9.686625325766818e-06, "loss": 0.0104, "step": 80860 }, { "grad_norm": 0.3055984377861023, "learning_rate": 9.67684641100554e-06, "loss": 0.0118, "step": 80870 }, { "grad_norm": 0.13860848546028137, "learning_rate": 9.667071905923692e-06, "loss": 0.0152, "step": 80880 }, { "grad_norm": 0.2625121772289276, "learning_rate": 9.657301811590242e-06, "loss": 0.0144, "step": 80890 }, { "grad_norm": 0.11289609223604202, "learning_rate": 9.647536129073603e-06, "loss": 0.0085, "step": 80900 }, { "grad_norm": 0.1560976654291153, "learning_rate": 9.637774859441751e-06, "loss": 0.011, "step": 80910 }, { "grad_norm": 0.07922422140836716, "learning_rate": 9.628018003762152e-06, "loss": 0.0071, "step": 80920 }, { "grad_norm": 0.19746437668800354, "learning_rate": 9.618265563101813e-06, "loss": 0.0103, "step": 80930 }, { "grad_norm": 0.13480058312416077, "learning_rate": 9.608517538527211e-06, "loss": 0.0147, "step": 80940 }, { "grad_norm": 0.29258230328559875, "learning_rate": 9.598773931104431e-06, "loss": 0.0109, "step": 80950 }, { "grad_norm": 0.1405656784772873, "learning_rate": 9.589034741898983e-06, "loss": 0.0124, "step": 80960 }, { "grad_norm": 0.06863205134868622, "learning_rate": 9.579299971975947e-06, "loss": 0.0077, "step": 80970 }, { "grad_norm": 0.1417539417743683, "learning_rate": 9.569569622399882e-06, "loss": 0.0124, "step": 80980 }, { "grad_norm": 0.08510706573724747, "learning_rate": 9.559843694234916e-06, "loss": 0.0088, "step": 80990 }, { "grad_norm": 0.12887825071811676, "learning_rate": 9.550122188544625e-06, "loss": 0.0101, "step": 81000 }, { "grad_norm": 0.08389285206794739, "learning_rate": 9.540405106392152e-06, "loss": 0.0117, "step": 81010 }, { "grad_norm": 0.0657806247472763, "learning_rate": 9.53069244884015e-06, "loss": 0.0109, "step": 81020 }, { "grad_norm": 0.07388227432966232, "learning_rate": 9.52098421695079e-06, "loss": 0.013, "step": 81030 }, { "grad_norm": 0.19653993844985962, "learning_rate": 9.511280411785728e-06, "loss": 0.0095, "step": 81040 }, { "grad_norm": 0.17246964573860168, "learning_rate": 9.50158103440616e-06, "loss": 0.0133, "step": 81050 }, { "grad_norm": 0.12298010289669037, "learning_rate": 9.491886085872815e-06, "loss": 0.0125, "step": 81060 }, { "grad_norm": 0.08847472816705704, "learning_rate": 9.482195567245895e-06, "loss": 0.0111, "step": 81070 }, { "grad_norm": 0.15872380137443542, "learning_rate": 9.472509479585151e-06, "loss": 0.0101, "step": 81080 }, { "grad_norm": 0.08279780298471451, "learning_rate": 9.462827823949833e-06, "loss": 0.0073, "step": 81090 }, { "grad_norm": 0.11400246620178223, "learning_rate": 9.453150601398735e-06, "loss": 0.0119, "step": 81100 }, { "grad_norm": 0.15555214881896973, "learning_rate": 9.44347781299011e-06, "loss": 0.0102, "step": 81110 }, { "grad_norm": 0.11906520277261734, "learning_rate": 9.43380945978179e-06, "loss": 0.0131, "step": 81120 }, { "grad_norm": 0.17497780919075012, "learning_rate": 9.424145542831047e-06, "loss": 0.0098, "step": 81130 }, { "grad_norm": 0.15000616014003754, "learning_rate": 9.414486063194778e-06, "loss": 0.0075, "step": 81140 }, { "grad_norm": 0.21359698474407196, "learning_rate": 9.404831021929273e-06, "loss": 0.0129, "step": 81150 }, { "grad_norm": 0.23858200013637543, "learning_rate": 9.395180420090427e-06, "loss": 0.0113, "step": 81160 }, { "grad_norm": 0.1258600503206253, "learning_rate": 9.385534258733591e-06, "loss": 0.0159, "step": 81170 }, { "grad_norm": 0.2530337870121002, "learning_rate": 9.375892538913677e-06, "loss": 0.0163, "step": 81180 }, { "grad_norm": 0.0763426423072815, "learning_rate": 9.36625526168507e-06, "loss": 0.0083, "step": 81190 }, { "grad_norm": 0.2334502637386322, "learning_rate": 9.356622428101686e-06, "loss": 0.0076, "step": 81200 }, { "grad_norm": 0.2798537015914917, "learning_rate": 9.346994039216972e-06, "loss": 0.0091, "step": 81210 }, { "grad_norm": 0.10013304650783539, "learning_rate": 9.337370096083886e-06, "loss": 0.0067, "step": 81220 }, { "grad_norm": 0.085369773209095, "learning_rate": 9.327750599754854e-06, "loss": 0.0136, "step": 81230 }, { "grad_norm": 0.03772466629743576, "learning_rate": 9.318135551281875e-06, "loss": 0.0104, "step": 81240 }, { "grad_norm": 0.10321569442749023, "learning_rate": 9.308524951716424e-06, "loss": 0.0153, "step": 81250 }, { "grad_norm": 0.2788773477077484, "learning_rate": 9.298918802109497e-06, "loss": 0.0143, "step": 81260 }, { "grad_norm": 0.07728376984596252, "learning_rate": 9.289317103511625e-06, "loss": 0.0073, "step": 81270 }, { "grad_norm": 0.2756049633026123, "learning_rate": 9.279719856972835e-06, "loss": 0.0152, "step": 81280 }, { "grad_norm": 0.35892847180366516, "learning_rate": 9.270127063542645e-06, "loss": 0.014, "step": 81290 }, { "grad_norm": 0.31500929594039917, "learning_rate": 9.260538724270124e-06, "loss": 0.018, "step": 81300 }, { "grad_norm": 0.136997789144516, "learning_rate": 9.250954840203852e-06, "loss": 0.0082, "step": 81310 }, { "grad_norm": 0.2857031226158142, "learning_rate": 9.241375412391878e-06, "loss": 0.0175, "step": 81320 }, { "grad_norm": 0.29242411255836487, "learning_rate": 9.23180044188181e-06, "loss": 0.0154, "step": 81330 }, { "grad_norm": 0.36983340978622437, "learning_rate": 9.222229929720755e-06, "loss": 0.0125, "step": 81340 }, { "grad_norm": 0.1280735433101654, "learning_rate": 9.212663876955336e-06, "loss": 0.0124, "step": 81350 }, { "grad_norm": 0.20427069067955017, "learning_rate": 9.20310228463166e-06, "loss": 0.0235, "step": 81360 }, { "grad_norm": 0.4360356330871582, "learning_rate": 9.19354515379539e-06, "loss": 0.0095, "step": 81370 }, { "grad_norm": 0.09605872631072998, "learning_rate": 9.183992485491666e-06, "loss": 0.0109, "step": 81380 }, { "grad_norm": 0.18374663591384888, "learning_rate": 9.174444280765154e-06, "loss": 0.0107, "step": 81390 }, { "grad_norm": 0.06294496357440948, "learning_rate": 9.164900540660032e-06, "loss": 0.0133, "step": 81400 }, { "grad_norm": 0.21697649359703064, "learning_rate": 9.155361266220008e-06, "loss": 0.0093, "step": 81410 }, { "grad_norm": 0.11854101717472076, "learning_rate": 9.145826458488254e-06, "loss": 0.0108, "step": 81420 }, { "grad_norm": 0.5569065809249878, "learning_rate": 9.136296118507509e-06, "loss": 0.0147, "step": 81430 }, { "grad_norm": 0.250813364982605, "learning_rate": 9.126770247319966e-06, "loss": 0.0191, "step": 81440 }, { "grad_norm": 0.10156559199094772, "learning_rate": 9.11724884596738e-06, "loss": 0.0103, "step": 81450 }, { "grad_norm": 0.13460029661655426, "learning_rate": 9.107731915490996e-06, "loss": 0.0094, "step": 81460 }, { "grad_norm": 0.11685149371623993, "learning_rate": 9.098219456931579e-06, "loss": 0.0069, "step": 81470 }, { "grad_norm": 0.24640336632728577, "learning_rate": 9.088711471329376e-06, "loss": 0.0113, "step": 81480 }, { "grad_norm": 0.19842728972434998, "learning_rate": 9.079207959724184e-06, "loss": 0.0126, "step": 81490 }, { "grad_norm": 0.06035676226019859, "learning_rate": 9.069708923155279e-06, "loss": 0.0147, "step": 81500 }, { "grad_norm": 0.17293182015419006, "learning_rate": 9.060214362661468e-06, "loss": 0.0065, "step": 81510 }, { "grad_norm": 0.08426941186189651, "learning_rate": 9.050724279281064e-06, "loss": 0.0104, "step": 81520 }, { "grad_norm": 0.10687349736690521, "learning_rate": 9.041238674051894e-06, "loss": 0.0069, "step": 81530 }, { "grad_norm": 0.09388699382543564, "learning_rate": 9.031757548011277e-06, "loss": 0.0264, "step": 81540 }, { "grad_norm": 0.247968852519989, "learning_rate": 9.022280902196051e-06, "loss": 0.0112, "step": 81550 }, { "grad_norm": 0.16098381578922272, "learning_rate": 9.012808737642592e-06, "loss": 0.0117, "step": 81560 }, { "grad_norm": 0.26199981570243835, "learning_rate": 9.003341055386739e-06, "loss": 0.0123, "step": 81570 }, { "grad_norm": 0.08443397283554077, "learning_rate": 8.993877856463862e-06, "loss": 0.0106, "step": 81580 }, { "grad_norm": 0.1586836278438568, "learning_rate": 8.984419141908856e-06, "loss": 0.0169, "step": 81590 }, { "grad_norm": 0.10407456755638123, "learning_rate": 8.974964912756117e-06, "loss": 0.0115, "step": 81600 }, { "grad_norm": 0.15375682711601257, "learning_rate": 8.965515170039523e-06, "loss": 0.0066, "step": 81610 }, { "grad_norm": 0.07341109961271286, "learning_rate": 8.956069914792514e-06, "loss": 0.0136, "step": 81620 }, { "grad_norm": 0.19801849126815796, "learning_rate": 8.946629148047969e-06, "loss": 0.0158, "step": 81630 }, { "grad_norm": 0.23332668840885162, "learning_rate": 8.93719287083834e-06, "loss": 0.0196, "step": 81640 }, { "grad_norm": 0.33244651556015015, "learning_rate": 8.927761084195569e-06, "loss": 0.0125, "step": 81650 }, { "grad_norm": 0.12014176696538925, "learning_rate": 8.918333789151096e-06, "loss": 0.0096, "step": 81660 }, { "grad_norm": 0.16448132693767548, "learning_rate": 8.90891098673587e-06, "loss": 0.0106, "step": 81670 }, { "grad_norm": 0.219669371843338, "learning_rate": 8.89949267798037e-06, "loss": 0.0122, "step": 81680 }, { "grad_norm": 0.08780073374509811, "learning_rate": 8.890078863914542e-06, "loss": 0.0081, "step": 81690 }, { "grad_norm": 0.10038477182388306, "learning_rate": 8.880669545567894e-06, "loss": 0.0119, "step": 81700 }, { "grad_norm": 0.22457820177078247, "learning_rate": 8.871264723969381e-06, "loss": 0.0127, "step": 81710 }, { "grad_norm": 0.04722364619374275, "learning_rate": 8.86186440014754e-06, "loss": 0.0101, "step": 81720 }, { "grad_norm": 0.09140300005674362, "learning_rate": 8.852468575130351e-06, "loss": 0.0274, "step": 81730 }, { "grad_norm": 0.11559519171714783, "learning_rate": 8.843077249945347e-06, "loss": 0.0086, "step": 81740 }, { "grad_norm": 0.2569355368614197, "learning_rate": 8.833690425619523e-06, "loss": 0.0118, "step": 81750 }, { "grad_norm": 0.18538323044776917, "learning_rate": 8.82430810317942e-06, "loss": 0.0189, "step": 81760 }, { "grad_norm": 0.23718421161174774, "learning_rate": 8.814930283651074e-06, "loss": 0.0103, "step": 81770 }, { "grad_norm": 0.091184601187706, "learning_rate": 8.805556968060048e-06, "loss": 0.0063, "step": 81780 }, { "grad_norm": 0.07493796944618225, "learning_rate": 8.796188157431362e-06, "loss": 0.012, "step": 81790 }, { "grad_norm": 0.1945916712284088, "learning_rate": 8.786823852789595e-06, "loss": 0.016, "step": 81800 }, { "grad_norm": 0.09109679609537125, "learning_rate": 8.777464055158813e-06, "loss": 0.0134, "step": 81810 }, { "grad_norm": 0.06235969811677933, "learning_rate": 8.768108765562578e-06, "loss": 0.0124, "step": 81820 }, { "grad_norm": 0.285165935754776, "learning_rate": 8.758757985023975e-06, "loss": 0.0119, "step": 81830 }, { "grad_norm": 0.11529112607240677, "learning_rate": 8.749411714565598e-06, "loss": 0.0078, "step": 81840 }, { "grad_norm": 0.035715408623218536, "learning_rate": 8.74006995520954e-06, "loss": 0.0136, "step": 81850 }, { "grad_norm": 0.1909756362438202, "learning_rate": 8.730732707977397e-06, "loss": 0.0089, "step": 81860 }, { "grad_norm": 0.15065467357635498, "learning_rate": 8.721399973890281e-06, "loss": 0.0101, "step": 81870 }, { "grad_norm": 0.2814828157424927, "learning_rate": 8.712071753968792e-06, "loss": 0.0182, "step": 81880 }, { "grad_norm": 0.1967935562133789, "learning_rate": 8.702748049233072e-06, "loss": 0.0161, "step": 81890 }, { "grad_norm": 0.12297537922859192, "learning_rate": 8.693428860702718e-06, "loss": 0.0119, "step": 81900 }, { "grad_norm": 0.065217986702919, "learning_rate": 8.684114189396898e-06, "loss": 0.0153, "step": 81910 }, { "grad_norm": 0.1174655333161354, "learning_rate": 8.674804036334222e-06, "loss": 0.0107, "step": 81920 }, { "grad_norm": 0.04554875195026398, "learning_rate": 8.665498402532857e-06, "loss": 0.0117, "step": 81930 }, { "grad_norm": 0.09175536781549454, "learning_rate": 8.656197289010432e-06, "loss": 0.0152, "step": 81940 }, { "grad_norm": 0.15427343547344208, "learning_rate": 8.646900696784122e-06, "loss": 0.0111, "step": 81950 }, { "grad_norm": 0.12152886390686035, "learning_rate": 8.637608626870553e-06, "loss": 0.0155, "step": 81960 }, { "grad_norm": 0.07468722760677338, "learning_rate": 8.628321080285944e-06, "loss": 0.0095, "step": 81970 }, { "grad_norm": 0.190302774310112, "learning_rate": 8.619038058045925e-06, "loss": 0.0114, "step": 81980 }, { "grad_norm": 0.13612446188926697, "learning_rate": 8.609759561165704e-06, "loss": 0.0095, "step": 81990 }, { "grad_norm": 0.10930067300796509, "learning_rate": 8.600485590659934e-06, "loss": 0.0094, "step": 82000 }, { "grad_norm": 0.0989559143781662, "learning_rate": 8.591216147542813e-06, "loss": 0.0137, "step": 82010 }, { "grad_norm": 0.04492951184511185, "learning_rate": 8.581951232828045e-06, "loss": 0.0259, "step": 82020 }, { "grad_norm": 0.14861701428890228, "learning_rate": 8.572690847528825e-06, "loss": 0.014, "step": 82030 }, { "grad_norm": 0.10789920389652252, "learning_rate": 8.563434992657838e-06, "loss": 0.0083, "step": 82040 }, { "grad_norm": 0.39875006675720215, "learning_rate": 8.554183669227307e-06, "loss": 0.0184, "step": 82050 }, { "grad_norm": 0.10817594826221466, "learning_rate": 8.544936878248937e-06, "loss": 0.0078, "step": 82060 }, { "grad_norm": 0.14676664769649506, "learning_rate": 8.535694620733942e-06, "loss": 0.0188, "step": 82070 }, { "grad_norm": 0.17378224432468414, "learning_rate": 8.526456897693053e-06, "loss": 0.0121, "step": 82080 }, { "grad_norm": 0.23409953713417053, "learning_rate": 8.517223710136456e-06, "loss": 0.0116, "step": 82090 }, { "grad_norm": 0.16711023449897766, "learning_rate": 8.507995059073931e-06, "loss": 0.0093, "step": 82100 }, { "grad_norm": 0.09190148860216141, "learning_rate": 8.498770945514673e-06, "loss": 0.0123, "step": 82110 }, { "grad_norm": 0.1087394654750824, "learning_rate": 8.489551370467436e-06, "loss": 0.0097, "step": 82120 }, { "grad_norm": 0.1990383267402649, "learning_rate": 8.480336334940442e-06, "loss": 0.016, "step": 82130 }, { "grad_norm": 0.10094178467988968, "learning_rate": 8.471125839941458e-06, "loss": 0.009, "step": 82140 }, { "grad_norm": 0.11289778351783752, "learning_rate": 8.461919886477682e-06, "loss": 0.012, "step": 82150 }, { "grad_norm": 0.06831181794404984, "learning_rate": 8.452718475555927e-06, "loss": 0.0078, "step": 82160 }, { "grad_norm": 0.1337081342935562, "learning_rate": 8.443521608182398e-06, "loss": 0.0124, "step": 82170 }, { "grad_norm": 0.15862241387367249, "learning_rate": 8.43432928536288e-06, "loss": 0.0107, "step": 82180 }, { "grad_norm": 0.10054602473974228, "learning_rate": 8.425141508102607e-06, "loss": 0.0126, "step": 82190 }, { "grad_norm": 0.17211250960826874, "learning_rate": 8.415958277406365e-06, "loss": 0.0142, "step": 82200 }, { "grad_norm": 0.0869186669588089, "learning_rate": 8.406779594278392e-06, "loss": 0.012, "step": 82210 }, { "grad_norm": 0.06293762475252151, "learning_rate": 8.397605459722468e-06, "loss": 0.0159, "step": 82220 }, { "grad_norm": 0.21195413172245026, "learning_rate": 8.388435874741863e-06, "loss": 0.0137, "step": 82230 }, { "grad_norm": 0.21703019738197327, "learning_rate": 8.37927084033936e-06, "loss": 0.0159, "step": 82240 }, { "grad_norm": 0.22391948103904724, "learning_rate": 8.370110357517208e-06, "loss": 0.0156, "step": 82250 }, { "grad_norm": 0.0739927738904953, "learning_rate": 8.360954427277201e-06, "loss": 0.0089, "step": 82260 }, { "grad_norm": 0.16112636029720306, "learning_rate": 8.351803050620627e-06, "loss": 0.0095, "step": 82270 }, { "grad_norm": 0.2785365581512451, "learning_rate": 8.34265622854824e-06, "loss": 0.019, "step": 82280 }, { "grad_norm": 0.0833173617720604, "learning_rate": 8.333513962060335e-06, "loss": 0.0112, "step": 82290 }, { "grad_norm": 0.07630191743373871, "learning_rate": 8.324376252156701e-06, "loss": 0.0101, "step": 82300 }, { "grad_norm": 0.16647090017795563, "learning_rate": 8.315243099836633e-06, "loss": 0.0146, "step": 82310 }, { "grad_norm": 0.0748453140258789, "learning_rate": 8.3061145060989e-06, "loss": 0.0173, "step": 82320 }, { "grad_norm": 0.3521818220615387, "learning_rate": 8.296990471941806e-06, "loss": 0.0127, "step": 82330 }, { "grad_norm": 0.10744509845972061, "learning_rate": 8.287870998363112e-06, "loss": 0.0133, "step": 82340 }, { "grad_norm": 0.22645023465156555, "learning_rate": 8.278756086360156e-06, "loss": 0.0109, "step": 82350 }, { "grad_norm": 0.14497452974319458, "learning_rate": 8.269645736929693e-06, "loss": 0.011, "step": 82360 }, { "grad_norm": 0.07005821913480759, "learning_rate": 8.260539951068048e-06, "loss": 0.0125, "step": 82370 }, { "grad_norm": 0.2359762340784073, "learning_rate": 8.251438729770983e-06, "loss": 0.0138, "step": 82380 }, { "grad_norm": 0.060578249394893646, "learning_rate": 8.242342074033826e-06, "loss": 0.0074, "step": 82390 }, { "grad_norm": 0.23483195900917053, "learning_rate": 8.233249984851344e-06, "loss": 0.0141, "step": 82400 }, { "grad_norm": 0.2840953469276428, "learning_rate": 8.224162463217849e-06, "loss": 0.0171, "step": 82410 }, { "grad_norm": 0.12946093082427979, "learning_rate": 8.215079510127139e-06, "loss": 0.0135, "step": 82420 }, { "grad_norm": 0.1655561774969101, "learning_rate": 8.206001126572521e-06, "loss": 0.0085, "step": 82430 }, { "grad_norm": 0.20353831350803375, "learning_rate": 8.196927313546776e-06, "loss": 0.0201, "step": 82440 }, { "grad_norm": 0.1846090704202652, "learning_rate": 8.18785807204222e-06, "loss": 0.0141, "step": 82450 }, { "grad_norm": 0.18684706091880798, "learning_rate": 8.178793403050627e-06, "loss": 0.0088, "step": 82460 }, { "grad_norm": 0.24152988195419312, "learning_rate": 8.16973330756331e-06, "loss": 0.0115, "step": 82470 }, { "grad_norm": 0.11879817396402359, "learning_rate": 8.16067778657107e-06, "loss": 0.0077, "step": 82480 }, { "grad_norm": 0.1362244337797165, "learning_rate": 8.151626841064214e-06, "loss": 0.0094, "step": 82490 }, { "grad_norm": 0.20228956639766693, "learning_rate": 8.142580472032518e-06, "loss": 0.0084, "step": 82500 }, { "grad_norm": 0.24049389362335205, "learning_rate": 8.133538680465285e-06, "loss": 0.0141, "step": 82510 }, { "grad_norm": 0.15162169933319092, "learning_rate": 8.124501467351336e-06, "loss": 0.0099, "step": 82520 }, { "grad_norm": 0.2135951668024063, "learning_rate": 8.11546883367893e-06, "loss": 0.0144, "step": 82530 }, { "grad_norm": 0.11189544945955276, "learning_rate": 8.106440780435882e-06, "loss": 0.0146, "step": 82540 }, { "grad_norm": 0.11895953863859177, "learning_rate": 8.097417308609485e-06, "loss": 0.0105, "step": 82550 }, { "grad_norm": 0.1116202101111412, "learning_rate": 8.088398419186544e-06, "loss": 0.0109, "step": 82560 }, { "grad_norm": 0.11834405362606049, "learning_rate": 8.079384113153321e-06, "loss": 0.0168, "step": 82570 }, { "grad_norm": 0.06648741662502289, "learning_rate": 8.070374391495644e-06, "loss": 0.0144, "step": 82580 }, { "grad_norm": 0.2553572356700897, "learning_rate": 8.061369255198764e-06, "loss": 0.0124, "step": 82590 }, { "grad_norm": 0.077947698533535, "learning_rate": 8.052368705247493e-06, "loss": 0.0146, "step": 82600 }, { "grad_norm": 0.06795214116573334, "learning_rate": 8.043372742626114e-06, "loss": 0.0187, "step": 82610 }, { "grad_norm": 0.24202974140644073, "learning_rate": 8.034381368318422e-06, "loss": 0.0111, "step": 82620 }, { "grad_norm": 0.29234975576400757, "learning_rate": 8.025394583307677e-06, "loss": 0.0126, "step": 82630 }, { "grad_norm": 0.3598335087299347, "learning_rate": 8.016412388576683e-06, "loss": 0.0158, "step": 82640 }, { "grad_norm": 0.1677524298429489, "learning_rate": 8.007434785107699e-06, "loss": 0.0078, "step": 82650 }, { "grad_norm": 0.17183861136436462, "learning_rate": 7.998461773882509e-06, "loss": 0.0097, "step": 82660 }, { "grad_norm": 0.15245600044727325, "learning_rate": 7.989493355882393e-06, "loss": 0.0073, "step": 82670 }, { "grad_norm": 0.15220244228839874, "learning_rate": 7.980529532088137e-06, "loss": 0.008, "step": 82680 }, { "grad_norm": 0.19512128829956055, "learning_rate": 7.97157030347998e-06, "loss": 0.0133, "step": 82690 }, { "grad_norm": 0.07230217754840851, "learning_rate": 7.96261567103772e-06, "loss": 0.0134, "step": 82700 }, { "grad_norm": 0.06045563519001007, "learning_rate": 7.953665635740593e-06, "loss": 0.0112, "step": 82710 }, { "grad_norm": 0.12593410909175873, "learning_rate": 7.94472019856738e-06, "loss": 0.0185, "step": 82720 }, { "grad_norm": 0.19879551231861115, "learning_rate": 7.935779360496337e-06, "loss": 0.012, "step": 82730 }, { "grad_norm": 0.2463020533323288, "learning_rate": 7.92684312250523e-06, "loss": 0.0155, "step": 82740 }, { "grad_norm": 0.18468715250492096, "learning_rate": 7.917911485571289e-06, "loss": 0.0146, "step": 82750 }, { "grad_norm": 0.2656826972961426, "learning_rate": 7.90898445067128e-06, "loss": 0.0109, "step": 82760 }, { "grad_norm": 0.13450011610984802, "learning_rate": 7.900062018781462e-06, "loss": 0.011, "step": 82770 }, { "grad_norm": 0.09013932943344116, "learning_rate": 7.891144190877553e-06, "loss": 0.0131, "step": 82780 }, { "grad_norm": 0.2779703140258789, "learning_rate": 7.882230967934806e-06, "loss": 0.0133, "step": 82790 }, { "grad_norm": 0.13744862377643585, "learning_rate": 7.873322350927959e-06, "loss": 0.0145, "step": 82800 }, { "grad_norm": 0.12266945838928223, "learning_rate": 7.864418340831253e-06, "loss": 0.0127, "step": 82810 }, { "grad_norm": 0.08199456334114075, "learning_rate": 7.855518938618395e-06, "loss": 0.009, "step": 82820 }, { "grad_norm": 0.12796035408973694, "learning_rate": 7.846624145262633e-06, "loss": 0.0118, "step": 82830 }, { "grad_norm": 0.09938249737024307, "learning_rate": 7.837733961736672e-06, "loss": 0.0161, "step": 82840 }, { "grad_norm": 0.111241914331913, "learning_rate": 7.828848389012727e-06, "loss": 0.0158, "step": 82850 }, { "grad_norm": 0.05147380381822586, "learning_rate": 7.819967428062524e-06, "loss": 0.0091, "step": 82860 }, { "grad_norm": 0.14642101526260376, "learning_rate": 7.811091079857274e-06, "loss": 0.0088, "step": 82870 }, { "grad_norm": 0.2331327646970749, "learning_rate": 7.802219345367667e-06, "loss": 0.0084, "step": 82880 }, { "grad_norm": 0.27597367763519287, "learning_rate": 7.793352225563916e-06, "loss": 0.0093, "step": 82890 }, { "grad_norm": 0.12591081857681274, "learning_rate": 7.7844897214157e-06, "loss": 0.0096, "step": 82900 }, { "grad_norm": 0.1681971549987793, "learning_rate": 7.775631833892228e-06, "loss": 0.0111, "step": 82910 }, { "grad_norm": 0.25277385115623474, "learning_rate": 7.76677856396215e-06, "loss": 0.0117, "step": 82920 }, { "grad_norm": 0.06451067328453064, "learning_rate": 7.757929912593697e-06, "loss": 0.008, "step": 82930 }, { "grad_norm": 0.1167004257440567, "learning_rate": 7.749085880754509e-06, "loss": 0.0113, "step": 82940 }, { "grad_norm": 0.2944515645503998, "learning_rate": 7.740246469411777e-06, "loss": 0.0116, "step": 82950 }, { "grad_norm": 0.29238152503967285, "learning_rate": 7.731411679532142e-06, "loss": 0.0104, "step": 82960 }, { "grad_norm": 0.14184997975826263, "learning_rate": 7.722581512081779e-06, "loss": 0.009, "step": 82970 }, { "grad_norm": 0.1712789237499237, "learning_rate": 7.713755968026331e-06, "loss": 0.0088, "step": 82980 }, { "grad_norm": 0.07097062468528748, "learning_rate": 7.704935048330969e-06, "loss": 0.0148, "step": 82990 }, { "grad_norm": 0.39584970474243164, "learning_rate": 7.696118753960313e-06, "loss": 0.0122, "step": 83000 }, { "grad_norm": 0.0766708254814148, "learning_rate": 7.687307085878503e-06, "loss": 0.0118, "step": 83010 }, { "grad_norm": 0.20811212062835693, "learning_rate": 7.678500045049186e-06, "loss": 0.0088, "step": 83020 }, { "grad_norm": 0.0792326107621193, "learning_rate": 7.669697632435463e-06, "loss": 0.0101, "step": 83030 }, { "grad_norm": 0.1903470903635025, "learning_rate": 7.660899848999963e-06, "loss": 0.0127, "step": 83040 }, { "grad_norm": 0.1681392788887024, "learning_rate": 7.652106695704803e-06, "loss": 0.0195, "step": 83050 }, { "grad_norm": 0.29442381858825684, "learning_rate": 7.643318173511598e-06, "loss": 0.013, "step": 83060 }, { "grad_norm": 0.15903958678245544, "learning_rate": 7.634534283381423e-06, "loss": 0.0131, "step": 83070 }, { "grad_norm": 0.5120337009429932, "learning_rate": 7.625755026274889e-06, "loss": 0.0146, "step": 83080 }, { "grad_norm": 0.15895937383174896, "learning_rate": 7.616980403152069e-06, "loss": 0.0093, "step": 83090 }, { "grad_norm": 0.07616093754768372, "learning_rate": 7.608210414972561e-06, "loss": 0.0167, "step": 83100 }, { "grad_norm": 0.13027611374855042, "learning_rate": 7.599445062695404e-06, "loss": 0.0153, "step": 83110 }, { "grad_norm": 0.09103995561599731, "learning_rate": 7.590684347279214e-06, "loss": 0.0119, "step": 83120 }, { "grad_norm": 0.09463828057050705, "learning_rate": 7.581928269682004e-06, "loss": 0.0092, "step": 83130 }, { "grad_norm": 0.5107762813568115, "learning_rate": 7.573176830861361e-06, "loss": 0.0108, "step": 83140 }, { "grad_norm": 0.28221774101257324, "learning_rate": 7.5644300317743e-06, "loss": 0.011, "step": 83150 }, { "grad_norm": 0.08531010150909424, "learning_rate": 7.555687873377376e-06, "loss": 0.0085, "step": 83160 }, { "grad_norm": 0.06746922433376312, "learning_rate": 7.546950356626592e-06, "loss": 0.0155, "step": 83170 }, { "grad_norm": 0.29273930191993713, "learning_rate": 7.538217482477516e-06, "loss": 0.011, "step": 83180 }, { "grad_norm": 0.08450191468000412, "learning_rate": 7.529489251885119e-06, "loss": 0.0109, "step": 83190 }, { "grad_norm": 0.12692423164844513, "learning_rate": 7.520765665803941e-06, "loss": 0.0092, "step": 83200 }, { "grad_norm": 0.43926581740379333, "learning_rate": 7.512046725187949e-06, "loss": 0.0116, "step": 83210 }, { "grad_norm": 0.07288388907909393, "learning_rate": 7.503332430990645e-06, "loss": 0.0126, "step": 83220 }, { "grad_norm": 0.37539392709732056, "learning_rate": 7.494622784165017e-06, "loss": 0.0157, "step": 83230 }, { "grad_norm": 0.23577088117599487, "learning_rate": 7.485917785663543e-06, "loss": 0.0105, "step": 83240 }, { "grad_norm": 0.2736680209636688, "learning_rate": 7.477217436438172e-06, "loss": 0.0124, "step": 83250 }, { "grad_norm": 0.16494528949260712, "learning_rate": 7.46852173744037e-06, "loss": 0.0108, "step": 83260 }, { "grad_norm": 0.18331734836101532, "learning_rate": 7.4598306896211004e-06, "loss": 0.0168, "step": 83270 }, { "grad_norm": 0.07831870764493942, "learning_rate": 7.451144293930773e-06, "loss": 0.0071, "step": 83280 }, { "grad_norm": 0.1762412190437317, "learning_rate": 7.4424625513193455e-06, "loss": 0.0086, "step": 83290 }, { "grad_norm": 0.3659347593784332, "learning_rate": 7.433785462736209e-06, "loss": 0.0131, "step": 83300 }, { "grad_norm": 0.21958135068416595, "learning_rate": 7.425113029130315e-06, "loss": 0.0132, "step": 83310 }, { "grad_norm": 0.13513125479221344, "learning_rate": 7.416445251450044e-06, "loss": 0.0163, "step": 83320 }, { "grad_norm": 0.13001416623592377, "learning_rate": 7.4077821306433035e-06, "loss": 0.009, "step": 83330 }, { "grad_norm": 0.048148199915885925, "learning_rate": 7.399123667657465e-06, "loss": 0.0073, "step": 83340 }, { "grad_norm": 0.13245484232902527, "learning_rate": 7.390469863439426e-06, "loss": 0.0111, "step": 83350 }, { "grad_norm": 0.12213099747896194, "learning_rate": 7.381820718935512e-06, "loss": 0.0103, "step": 83360 }, { "grad_norm": 0.26659706234931946, "learning_rate": 7.373176235091633e-06, "loss": 0.0156, "step": 83370 }, { "grad_norm": 0.1241912692785263, "learning_rate": 7.364536412853101e-06, "loss": 0.0129, "step": 83380 }, { "grad_norm": 0.1416759043931961, "learning_rate": 7.355901253164776e-06, "loss": 0.0091, "step": 83390 }, { "grad_norm": 0.07379503548145294, "learning_rate": 7.34727075697097e-06, "loss": 0.0093, "step": 83400 }, { "grad_norm": 0.13193412125110626, "learning_rate": 7.3386449252155095e-06, "loss": 0.0073, "step": 83410 }, { "grad_norm": 0.06884795427322388, "learning_rate": 7.330023758841692e-06, "loss": 0.0181, "step": 83420 }, { "grad_norm": 0.051076579838991165, "learning_rate": 7.321407258792323e-06, "loss": 0.0044, "step": 83430 }, { "grad_norm": 0.08275462687015533, "learning_rate": 7.312795426009694e-06, "loss": 0.007, "step": 83440 }, { "grad_norm": 0.1667180359363556, "learning_rate": 7.304188261435585e-06, "loss": 0.009, "step": 83450 }, { "grad_norm": 0.1602349728345871, "learning_rate": 7.295585766011242e-06, "loss": 0.0136, "step": 83460 }, { "grad_norm": 0.03751341253519058, "learning_rate": 7.286987940677436e-06, "loss": 0.0149, "step": 83470 }, { "grad_norm": 0.15440870821475983, "learning_rate": 7.278394786374426e-06, "loss": 0.0055, "step": 83480 }, { "grad_norm": 0.11187354475259781, "learning_rate": 7.269806304041915e-06, "loss": 0.0094, "step": 83490 }, { "grad_norm": 0.080852210521698, "learning_rate": 7.261222494619147e-06, "loss": 0.0091, "step": 83500 }, { "grad_norm": 0.18311730027198792, "learning_rate": 7.252643359044825e-06, "loss": 0.0126, "step": 83510 }, { "grad_norm": 0.4196097254753113, "learning_rate": 7.244068898257172e-06, "loss": 0.0111, "step": 83520 }, { "grad_norm": 0.27475759387016296, "learning_rate": 7.2354991131938495e-06, "loss": 0.0073, "step": 83530 }, { "grad_norm": 0.07928922772407532, "learning_rate": 7.2269340047920574e-06, "loss": 0.0141, "step": 83540 }, { "grad_norm": 0.18437567353248596, "learning_rate": 7.218373573988435e-06, "loss": 0.0072, "step": 83550 }, { "grad_norm": 0.2016335278749466, "learning_rate": 7.209817821719178e-06, "loss": 0.0125, "step": 83560 }, { "grad_norm": 0.2934924066066742, "learning_rate": 7.201266748919894e-06, "loss": 0.0136, "step": 83570 }, { "grad_norm": 0.2319566309452057, "learning_rate": 7.19272035652575e-06, "loss": 0.0107, "step": 83580 }, { "grad_norm": 0.2757597863674164, "learning_rate": 7.184178645471334e-06, "loss": 0.0101, "step": 83590 }, { "grad_norm": 0.0821896344423294, "learning_rate": 7.175641616690775e-06, "loss": 0.0092, "step": 83600 }, { "grad_norm": 0.35204175114631653, "learning_rate": 7.167109271117661e-06, "loss": 0.0144, "step": 83610 }, { "grad_norm": 0.13842524588108063, "learning_rate": 7.158581609685072e-06, "loss": 0.0154, "step": 83620 }, { "grad_norm": 0.12491012364625931, "learning_rate": 7.150058633325591e-06, "loss": 0.009, "step": 83630 }, { "grad_norm": 0.11542581021785736, "learning_rate": 7.141540342971287e-06, "loss": 0.0114, "step": 83640 }, { "grad_norm": 0.15989194810390472, "learning_rate": 7.133026739553677e-06, "loss": 0.009, "step": 83650 }, { "grad_norm": 0.16586972773075104, "learning_rate": 7.124517824003829e-06, "loss": 0.0096, "step": 83660 }, { "grad_norm": 0.2097589075565338, "learning_rate": 7.116013597252236e-06, "loss": 0.0161, "step": 83670 }, { "grad_norm": 0.3990010619163513, "learning_rate": 7.107514060228921e-06, "loss": 0.0158, "step": 83680 }, { "grad_norm": 0.2807314395904541, "learning_rate": 7.099019213863378e-06, "loss": 0.0164, "step": 83690 }, { "grad_norm": 0.2318895161151886, "learning_rate": 7.090529059084611e-06, "loss": 0.0121, "step": 83700 }, { "grad_norm": 0.06790429353713989, "learning_rate": 7.082043596821058e-06, "loss": 0.0063, "step": 83710 }, { "grad_norm": 0.1518615037202835, "learning_rate": 7.07356282800069e-06, "loss": 0.0068, "step": 83720 }, { "grad_norm": 0.08611205965280533, "learning_rate": 7.0650867535509645e-06, "loss": 0.0123, "step": 83730 }, { "grad_norm": 0.23125964403152466, "learning_rate": 7.0566153743987895e-06, "loss": 0.0115, "step": 83740 }, { "grad_norm": 0.2863336205482483, "learning_rate": 7.04814869147059e-06, "loss": 0.0165, "step": 83750 }, { "grad_norm": 0.2839549481868744, "learning_rate": 7.0396867056922725e-06, "loss": 0.015, "step": 83760 }, { "grad_norm": 0.20093418657779694, "learning_rate": 7.031229417989244e-06, "loss": 0.0135, "step": 83770 }, { "grad_norm": 0.10507381707429886, "learning_rate": 7.022776829286348e-06, "loss": 0.0092, "step": 83780 }, { "grad_norm": 0.07122456282377243, "learning_rate": 7.014328940507975e-06, "loss": 0.0087, "step": 83790 }, { "grad_norm": 0.2728137671947479, "learning_rate": 7.005885752577945e-06, "loss": 0.0128, "step": 83800 }, { "grad_norm": 0.275642454624176, "learning_rate": 6.997447266419615e-06, "loss": 0.0127, "step": 83810 }, { "grad_norm": 0.19310720264911652, "learning_rate": 6.989013482955797e-06, "loss": 0.0109, "step": 83820 }, { "grad_norm": 0.1520644724369049, "learning_rate": 6.980584403108808e-06, "loss": 0.0123, "step": 83830 }, { "grad_norm": 0.2343655526638031, "learning_rate": 6.972160027800417e-06, "loss": 0.0112, "step": 83840 }, { "grad_norm": 0.2636833190917969, "learning_rate": 6.963740357951931e-06, "loss": 0.0142, "step": 83850 }, { "grad_norm": 0.2323947697877884, "learning_rate": 6.95532539448408e-06, "loss": 0.0086, "step": 83860 }, { "grad_norm": 0.14146366715431213, "learning_rate": 6.946915138317128e-06, "loss": 0.0104, "step": 83870 }, { "grad_norm": 0.2775361239910126, "learning_rate": 6.9385095903708065e-06, "loss": 0.0124, "step": 83880 }, { "grad_norm": 0.17197047173976898, "learning_rate": 6.930108751564346e-06, "loss": 0.0117, "step": 83890 }, { "grad_norm": 0.566568911075592, "learning_rate": 6.921712622816428e-06, "loss": 0.0169, "step": 83900 }, { "grad_norm": 0.11386998742818832, "learning_rate": 6.913321205045259e-06, "loss": 0.0088, "step": 83910 }, { "grad_norm": 0.28107815980911255, "learning_rate": 6.904934499168491e-06, "loss": 0.0129, "step": 83920 }, { "grad_norm": 0.1467117965221405, "learning_rate": 6.896552506103299e-06, "loss": 0.0109, "step": 83930 }, { "grad_norm": 0.21589893102645874, "learning_rate": 6.888175226766313e-06, "loss": 0.0149, "step": 83940 }, { "grad_norm": 0.34811151027679443, "learning_rate": 6.879802662073686e-06, "loss": 0.0137, "step": 83950 }, { "grad_norm": 0.33150801062583923, "learning_rate": 6.871434812940991e-06, "loss": 0.0075, "step": 83960 }, { "grad_norm": 0.19899289309978485, "learning_rate": 6.863071680283345e-06, "loss": 0.0076, "step": 83970 }, { "grad_norm": 0.11392639577388763, "learning_rate": 6.854713265015339e-06, "loss": 0.0299, "step": 83980 }, { "grad_norm": 0.22492466866970062, "learning_rate": 6.8463595680510025e-06, "loss": 0.0157, "step": 83990 }, { "grad_norm": 0.2723426818847656, "learning_rate": 6.838010590303911e-06, "loss": 0.0095, "step": 84000 }, { "grad_norm": 0.1515466272830963, "learning_rate": 6.829666332687085e-06, "loss": 0.0087, "step": 84010 }, { "grad_norm": 0.05542986840009689, "learning_rate": 6.821326796113059e-06, "loss": 0.0106, "step": 84020 }, { "grad_norm": 0.2811793088912964, "learning_rate": 6.812991981493799e-06, "loss": 0.009, "step": 84030 }, { "grad_norm": 0.02990974672138691, "learning_rate": 6.804661889740821e-06, "loss": 0.0124, "step": 84040 }, { "grad_norm": 0.06362847238779068, "learning_rate": 6.7963365217650625e-06, "loss": 0.0103, "step": 84050 }, { "grad_norm": 0.08097178488969803, "learning_rate": 6.788015878476983e-06, "loss": 0.0156, "step": 84060 }, { "grad_norm": 0.3052343428134918, "learning_rate": 6.779699960786523e-06, "loss": 0.0106, "step": 84070 }, { "grad_norm": 0.0727384015917778, "learning_rate": 6.771388769603104e-06, "loss": 0.0127, "step": 84080 }, { "grad_norm": 0.1643429696559906, "learning_rate": 6.763082305835605e-06, "loss": 0.0239, "step": 84090 }, { "grad_norm": 0.14174199104309082, "learning_rate": 6.754780570392433e-06, "loss": 0.0089, "step": 84100 }, { "grad_norm": 0.17690420150756836, "learning_rate": 6.7464835641814215e-06, "loss": 0.0119, "step": 84110 }, { "grad_norm": 0.32514894008636475, "learning_rate": 6.7381912881099515e-06, "loss": 0.0105, "step": 84120 }, { "grad_norm": 0.07226446270942688, "learning_rate": 6.72990374308482e-06, "loss": 0.0096, "step": 84130 }, { "grad_norm": 0.09800711274147034, "learning_rate": 6.721620930012379e-06, "loss": 0.0107, "step": 84140 }, { "grad_norm": 0.07778120785951614, "learning_rate": 6.713342849798393e-06, "loss": 0.0068, "step": 84150 }, { "grad_norm": 0.09638948738574982, "learning_rate": 6.705069503348166e-06, "loss": 0.0077, "step": 84160 }, { "grad_norm": 0.301142156124115, "learning_rate": 6.696800891566435e-06, "loss": 0.011, "step": 84170 }, { "grad_norm": 0.21699051558971405, "learning_rate": 6.688537015357449e-06, "loss": 0.0105, "step": 84180 }, { "grad_norm": 0.08304579555988312, "learning_rate": 6.680277875624941e-06, "loss": 0.0125, "step": 84190 }, { "grad_norm": 0.06343917548656464, "learning_rate": 6.672023473272121e-06, "loss": 0.0135, "step": 84200 }, { "grad_norm": 0.09128701686859131, "learning_rate": 6.663773809201667e-06, "loss": 0.0112, "step": 84210 }, { "grad_norm": 0.050528015941381454, "learning_rate": 6.655528884315748e-06, "loss": 0.0083, "step": 84220 }, { "grad_norm": 0.17083311080932617, "learning_rate": 6.647288699516036e-06, "loss": 0.0149, "step": 84230 }, { "grad_norm": 0.1595081239938736, "learning_rate": 6.639053255703642e-06, "loss": 0.0177, "step": 84240 }, { "grad_norm": 0.2352796345949173, "learning_rate": 6.6308225537791925e-06, "loss": 0.0109, "step": 84250 }, { "grad_norm": 0.055036164820194244, "learning_rate": 6.622596594642777e-06, "loss": 0.0092, "step": 84260 }, { "grad_norm": 0.2848486304283142, "learning_rate": 6.614375379193999e-06, "loss": 0.0122, "step": 84270 }, { "grad_norm": 0.03822341561317444, "learning_rate": 6.606158908331889e-06, "loss": 0.0123, "step": 84280 }, { "grad_norm": 0.1624443233013153, "learning_rate": 6.597947182955005e-06, "loss": 0.009, "step": 84290 }, { "grad_norm": 0.16514228284358978, "learning_rate": 6.589740203961359e-06, "loss": 0.017, "step": 84300 }, { "grad_norm": 0.1511821299791336, "learning_rate": 6.581537972248464e-06, "loss": 0.0106, "step": 84310 }, { "grad_norm": 0.24165567755699158, "learning_rate": 6.573340488713275e-06, "loss": 0.0069, "step": 84320 }, { "grad_norm": 0.42696887254714966, "learning_rate": 6.565147754252304e-06, "loss": 0.0147, "step": 84330 }, { "grad_norm": 0.1027827262878418, "learning_rate": 6.556959769761462e-06, "loss": 0.008, "step": 84340 }, { "grad_norm": 0.16958844661712646, "learning_rate": 6.548776536136192e-06, "loss": 0.0078, "step": 84350 }, { "grad_norm": 0.2077866643667221, "learning_rate": 6.54059805427138e-06, "loss": 0.0101, "step": 84360 }, { "grad_norm": 0.0921863242983818, "learning_rate": 6.532424325061437e-06, "loss": 0.0181, "step": 84370 }, { "grad_norm": 0.11939480155706406, "learning_rate": 6.5242553494001935e-06, "loss": 0.0104, "step": 84380 }, { "grad_norm": 0.16265998780727386, "learning_rate": 6.516091128181046e-06, "loss": 0.0153, "step": 84390 }, { "grad_norm": 0.2473669946193695, "learning_rate": 6.507931662296785e-06, "loss": 0.0093, "step": 84400 }, { "grad_norm": 0.21075427532196045, "learning_rate": 6.4997769526397354e-06, "loss": 0.0087, "step": 84410 }, { "grad_norm": 0.26011380553245544, "learning_rate": 6.491627000101663e-06, "loss": 0.0105, "step": 84420 }, { "grad_norm": 0.15223939716815948, "learning_rate": 6.4834818055738545e-06, "loss": 0.0131, "step": 84430 }, { "grad_norm": 0.1186329796910286, "learning_rate": 6.4753413699470465e-06, "loss": 0.0089, "step": 84440 }, { "grad_norm": 0.3739371597766876, "learning_rate": 6.467205694111478e-06, "loss": 0.0177, "step": 84450 }, { "grad_norm": 0.17554239928722382, "learning_rate": 6.459074778956836e-06, "loss": 0.0145, "step": 84460 }, { "grad_norm": 0.0853550136089325, "learning_rate": 6.450948625372311e-06, "loss": 0.0127, "step": 84470 }, { "grad_norm": 0.044059865176677704, "learning_rate": 6.442827234246585e-06, "loss": 0.0071, "step": 84480 }, { "grad_norm": 0.08977489918470383, "learning_rate": 6.43471060646777e-06, "loss": 0.0093, "step": 84490 }, { "grad_norm": 0.10802639275789261, "learning_rate": 6.426598742923512e-06, "loss": 0.0089, "step": 84500 }, { "grad_norm": 0.5132333040237427, "learning_rate": 6.418491644500885e-06, "loss": 0.0244, "step": 84510 }, { "grad_norm": 0.1965043544769287, "learning_rate": 6.410389312086512e-06, "loss": 0.0151, "step": 84520 }, { "grad_norm": 0.06453631073236465, "learning_rate": 6.4022917465664135e-06, "loss": 0.0131, "step": 84530 }, { "grad_norm": 0.14040113985538483, "learning_rate": 6.3941989488261575e-06, "loss": 0.0086, "step": 84540 }, { "grad_norm": 0.22242873907089233, "learning_rate": 6.386110919750726e-06, "loss": 0.0122, "step": 84550 }, { "grad_norm": 0.13104921579360962, "learning_rate": 6.378027660224651e-06, "loss": 0.008, "step": 84560 }, { "grad_norm": 0.35707658529281616, "learning_rate": 6.369949171131856e-06, "loss": 0.0204, "step": 84570 }, { "grad_norm": 0.17169882357120514, "learning_rate": 6.361875453355848e-06, "loss": 0.0101, "step": 84580 }, { "grad_norm": 0.10227999091148376, "learning_rate": 6.353806507779525e-06, "loss": 0.0106, "step": 84590 }, { "grad_norm": 0.16733397543430328, "learning_rate": 6.3457423352853044e-06, "loss": 0.0097, "step": 84600 }, { "grad_norm": 0.042088333517313004, "learning_rate": 6.337682936755062e-06, "loss": 0.0142, "step": 84610 }, { "grad_norm": 0.24626700580120087, "learning_rate": 6.329628313070179e-06, "loss": 0.012, "step": 84620 }, { "grad_norm": 0.15032830834388733, "learning_rate": 6.321578465111477e-06, "loss": 0.0105, "step": 84630 }, { "grad_norm": 0.08696303516626358, "learning_rate": 6.31353339375928e-06, "loss": 0.0091, "step": 84640 }, { "grad_norm": 0.3110964000225067, "learning_rate": 6.305493099893395e-06, "loss": 0.0151, "step": 84650 }, { "grad_norm": 0.08757121860980988, "learning_rate": 6.297457584393096e-06, "loss": 0.0077, "step": 84660 }, { "grad_norm": 0.08007661253213882, "learning_rate": 6.289426848137126e-06, "loss": 0.0136, "step": 84670 }, { "grad_norm": 0.16406969726085663, "learning_rate": 6.281400892003713e-06, "loss": 0.0103, "step": 84680 }, { "grad_norm": 0.2419319450855255, "learning_rate": 6.273379716870581e-06, "loss": 0.0115, "step": 84690 }, { "grad_norm": 0.21120046079158783, "learning_rate": 6.265363323614887e-06, "loss": 0.0142, "step": 84700 }, { "grad_norm": 0.07000353187322617, "learning_rate": 6.257351713113308e-06, "loss": 0.0127, "step": 84710 }, { "grad_norm": 0.17685037851333618, "learning_rate": 6.249344886241981e-06, "loss": 0.0062, "step": 84720 }, { "grad_norm": 0.043733686208724976, "learning_rate": 6.2413428438765295e-06, "loss": 0.0076, "step": 84730 }, { "grad_norm": 0.35330772399902344, "learning_rate": 6.233345586892026e-06, "loss": 0.0081, "step": 84740 }, { "grad_norm": 0.34635376930236816, "learning_rate": 6.225353116163052e-06, "loss": 0.0135, "step": 84750 }, { "grad_norm": 0.07614617794752121, "learning_rate": 6.217365432563627e-06, "loss": 0.0107, "step": 84760 }, { "grad_norm": 0.18044477701187134, "learning_rate": 6.209382536967318e-06, "loss": 0.0095, "step": 84770 }, { "grad_norm": 0.18380357325077057, "learning_rate": 6.201404430247082e-06, "loss": 0.0108, "step": 84780 }, { "grad_norm": 0.2180507779121399, "learning_rate": 6.1934311132754165e-06, "loss": 0.015, "step": 84790 }, { "grad_norm": 0.11049504578113556, "learning_rate": 6.185462586924257e-06, "loss": 0.0128, "step": 84800 }, { "grad_norm": 0.2388995885848999, "learning_rate": 6.177498852065039e-06, "loss": 0.0153, "step": 84810 }, { "grad_norm": 0.07589007169008255, "learning_rate": 6.169539909568655e-06, "loss": 0.0159, "step": 84820 }, { "grad_norm": 0.06719907373189926, "learning_rate": 6.161585760305488e-06, "loss": 0.0123, "step": 84830 }, { "grad_norm": 0.1460397094488144, "learning_rate": 6.153636405145391e-06, "loss": 0.0096, "step": 84840 }, { "grad_norm": 0.19050733745098114, "learning_rate": 6.145691844957702e-06, "loss": 0.011, "step": 84850 }, { "grad_norm": 0.06219862774014473, "learning_rate": 6.1377520806112145e-06, "loss": 0.0105, "step": 84860 }, { "grad_norm": 0.1597670167684555, "learning_rate": 6.129817112974218e-06, "loss": 0.013, "step": 84870 }, { "grad_norm": 0.2526320517063141, "learning_rate": 6.121886942914457e-06, "loss": 0.0106, "step": 84880 }, { "grad_norm": 0.18040981888771057, "learning_rate": 6.11396157129917e-06, "loss": 0.0122, "step": 84890 }, { "grad_norm": 0.1781671941280365, "learning_rate": 6.106040998995066e-06, "loss": 0.0121, "step": 84900 }, { "grad_norm": 0.206670880317688, "learning_rate": 6.098125226868329e-06, "loss": 0.0122, "step": 84910 }, { "grad_norm": 0.35182899236679077, "learning_rate": 6.090214255784604e-06, "loss": 0.0192, "step": 84920 }, { "grad_norm": 0.33594605326652527, "learning_rate": 6.0823080866090346e-06, "loss": 0.02, "step": 84930 }, { "grad_norm": 0.14943882822990417, "learning_rate": 6.074406720206233e-06, "loss": 0.0089, "step": 84940 }, { "grad_norm": 0.7709595561027527, "learning_rate": 6.066510157440259e-06, "loss": 0.0146, "step": 84950 }, { "grad_norm": 0.0748843103647232, "learning_rate": 6.058618399174676e-06, "loss": 0.0111, "step": 84960 }, { "grad_norm": 0.10137901455163956, "learning_rate": 6.050731446272528e-06, "loss": 0.0114, "step": 84970 }, { "grad_norm": 0.10816141963005066, "learning_rate": 6.042849299596321e-06, "loss": 0.0074, "step": 84980 }, { "grad_norm": 0.11925702542066574, "learning_rate": 6.034971960008013e-06, "loss": 0.0121, "step": 84990 }, { "grad_norm": 0.04869915917515755, "learning_rate": 6.027099428369082e-06, "loss": 0.0125, "step": 85000 }, { "grad_norm": 0.401096373796463, "learning_rate": 6.019231705540435e-06, "loss": 0.0132, "step": 85010 }, { "grad_norm": 0.04796605184674263, "learning_rate": 6.011368792382488e-06, "loss": 0.0115, "step": 85020 }, { "grad_norm": 0.10105869919061661, "learning_rate": 6.00351068975511e-06, "loss": 0.01, "step": 85030 }, { "grad_norm": 0.05242004245519638, "learning_rate": 5.9956573985176674e-06, "loss": 0.0111, "step": 85040 }, { "grad_norm": 0.10298052430152893, "learning_rate": 5.987808919528959e-06, "loss": 0.0131, "step": 85050 }, { "grad_norm": 0.1220802292227745, "learning_rate": 5.979965253647307e-06, "loss": 0.0193, "step": 85060 }, { "grad_norm": 0.14721526205539703, "learning_rate": 5.9721264017304655e-06, "loss": 0.0116, "step": 85070 }, { "grad_norm": 0.10717856138944626, "learning_rate": 5.964292364635682e-06, "loss": 0.0091, "step": 85080 }, { "grad_norm": 0.04691385477781296, "learning_rate": 5.956463143219682e-06, "loss": 0.0134, "step": 85090 }, { "grad_norm": 0.17098744213581085, "learning_rate": 5.948638738338663e-06, "loss": 0.0137, "step": 85100 }, { "grad_norm": 0.03853468596935272, "learning_rate": 5.940819150848271e-06, "loss": 0.0105, "step": 85110 }, { "grad_norm": 0.21279801428318024, "learning_rate": 5.933004381603663e-06, "loss": 0.0117, "step": 85120 }, { "grad_norm": 0.16888484358787537, "learning_rate": 5.92519443145943e-06, "loss": 0.0091, "step": 85130 }, { "grad_norm": 0.4123508334159851, "learning_rate": 5.917389301269671e-06, "loss": 0.0166, "step": 85140 }, { "grad_norm": 0.12447846680879593, "learning_rate": 5.9095889918879435e-06, "loss": 0.0118, "step": 85150 }, { "grad_norm": 0.0974169448018074, "learning_rate": 5.901793504167286e-06, "loss": 0.0136, "step": 85160 }, { "grad_norm": 0.09492004662752151, "learning_rate": 5.894002838960172e-06, "loss": 0.0189, "step": 85170 }, { "grad_norm": 0.22507086396217346, "learning_rate": 5.886216997118604e-06, "loss": 0.014, "step": 85180 }, { "grad_norm": 0.23633964359760284, "learning_rate": 5.878435979494023e-06, "loss": 0.0127, "step": 85190 }, { "grad_norm": 0.057232052087783813, "learning_rate": 5.870659786937344e-06, "loss": 0.0078, "step": 85200 }, { "grad_norm": 0.05787632241845131, "learning_rate": 5.862888420298962e-06, "loss": 0.0064, "step": 85210 }, { "grad_norm": 0.1623051017522812, "learning_rate": 5.855121880428743e-06, "loss": 0.0079, "step": 85220 }, { "grad_norm": 0.3467988073825836, "learning_rate": 5.84736016817603e-06, "loss": 0.0111, "step": 85230 }, { "grad_norm": 0.1650800108909607, "learning_rate": 5.839603284389622e-06, "loss": 0.0082, "step": 85240 }, { "grad_norm": 0.14425452053546906, "learning_rate": 5.831851229917812e-06, "loss": 0.0102, "step": 85250 }, { "grad_norm": 0.13484564423561096, "learning_rate": 5.824104005608338e-06, "loss": 0.0118, "step": 85260 }, { "grad_norm": 0.07656116038560867, "learning_rate": 5.81636161230843e-06, "loss": 0.0152, "step": 85270 }, { "grad_norm": 0.22155705094337463, "learning_rate": 5.808624050864791e-06, "loss": 0.0133, "step": 85280 }, { "grad_norm": 0.1494012326002121, "learning_rate": 5.800891322123592e-06, "loss": 0.0128, "step": 85290 }, { "grad_norm": 0.10321668535470963, "learning_rate": 5.793163426930454e-06, "loss": 0.0111, "step": 85300 }, { "grad_norm": 0.4402763247489929, "learning_rate": 5.785440366130512e-06, "loss": 0.0113, "step": 85310 }, { "grad_norm": 0.05530771240592003, "learning_rate": 5.777722140568326e-06, "loss": 0.0048, "step": 85320 }, { "grad_norm": 0.055290449410676956, "learning_rate": 5.7700087510879685e-06, "loss": 0.0131, "step": 85330 }, { "grad_norm": 0.15122218430042267, "learning_rate": 5.762300198532933e-06, "loss": 0.0088, "step": 85340 }, { "grad_norm": 0.6050603985786438, "learning_rate": 5.7545964837462606e-06, "loss": 0.0082, "step": 85350 }, { "grad_norm": 0.11079486459493637, "learning_rate": 5.746897607570384e-06, "loss": 0.0072, "step": 85360 }, { "grad_norm": 0.04988476634025574, "learning_rate": 5.7392035708472626e-06, "loss": 0.0075, "step": 85370 }, { "grad_norm": 0.21096017956733704, "learning_rate": 5.731514374418278e-06, "loss": 0.0145, "step": 85380 }, { "grad_norm": 0.11432749778032303, "learning_rate": 5.7238300191243325e-06, "loss": 0.0145, "step": 85390 }, { "grad_norm": 0.13416603207588196, "learning_rate": 5.716150505805762e-06, "loss": 0.0156, "step": 85400 }, { "grad_norm": 0.09361919015645981, "learning_rate": 5.708475835302407e-06, "loss": 0.0093, "step": 85410 }, { "grad_norm": 0.20945395529270172, "learning_rate": 5.70080600845353e-06, "loss": 0.0211, "step": 85420 }, { "grad_norm": 0.2678759694099426, "learning_rate": 5.693141026097909e-06, "loss": 0.0155, "step": 85430 }, { "grad_norm": 0.08672995865345001, "learning_rate": 5.6854808890737745e-06, "loss": 0.0107, "step": 85440 }, { "grad_norm": 0.0695432722568512, "learning_rate": 5.6778255982188225e-06, "loss": 0.0132, "step": 85450 }, { "grad_norm": 0.13830037415027618, "learning_rate": 5.670175154370217e-06, "loss": 0.0096, "step": 85460 }, { "grad_norm": 0.0894811600446701, "learning_rate": 5.662529558364616e-06, "loss": 0.0076, "step": 85470 }, { "grad_norm": 0.276515394449234, "learning_rate": 5.654888811038128e-06, "loss": 0.013, "step": 85480 }, { "grad_norm": 0.10221455991268158, "learning_rate": 5.647252913226314e-06, "loss": 0.0094, "step": 85490 }, { "grad_norm": 0.3529386818408966, "learning_rate": 5.639621865764255e-06, "loss": 0.0136, "step": 85500 }, { "grad_norm": 0.3134745657444, "learning_rate": 5.63199566948644e-06, "loss": 0.0143, "step": 85510 }, { "grad_norm": 0.050784461200237274, "learning_rate": 5.6243743252268835e-06, "loss": 0.0129, "step": 85520 }, { "grad_norm": 0.22276918590068817, "learning_rate": 5.616757833819009e-06, "loss": 0.0102, "step": 85530 }, { "grad_norm": 0.20014487206935883, "learning_rate": 5.609146196095794e-06, "loss": 0.0101, "step": 85540 }, { "grad_norm": 0.2127103954553604, "learning_rate": 5.6015394128896005e-06, "loss": 0.009, "step": 85550 }, { "grad_norm": 0.16589593887329102, "learning_rate": 5.593937485032308e-06, "loss": 0.0137, "step": 85560 }, { "grad_norm": 0.25955814123153687, "learning_rate": 5.586340413355246e-06, "loss": 0.0154, "step": 85570 }, { "grad_norm": 0.13527323305606842, "learning_rate": 5.578748198689227e-06, "loss": 0.0136, "step": 85580 }, { "grad_norm": 0.2623746991157532, "learning_rate": 5.571160841864504e-06, "loss": 0.0129, "step": 85590 }, { "grad_norm": 0.13879744708538055, "learning_rate": 5.563578343710846e-06, "loss": 0.0046, "step": 85600 }, { "grad_norm": 0.2365121692419052, "learning_rate": 5.556000705057446e-06, "loss": 0.0085, "step": 85610 }, { "grad_norm": 0.11410627514123917, "learning_rate": 5.548427926732996e-06, "loss": 0.0086, "step": 85620 }, { "grad_norm": 0.12532371282577515, "learning_rate": 5.540860009565629e-06, "loss": 0.0063, "step": 85630 }, { "grad_norm": 0.1539204865694046, "learning_rate": 5.533296954382966e-06, "loss": 0.0121, "step": 85640 }, { "grad_norm": 0.14319390058517456, "learning_rate": 5.525738762012095e-06, "loss": 0.0128, "step": 85650 }, { "grad_norm": 0.09494418650865555, "learning_rate": 5.518185433279577e-06, "loss": 0.0102, "step": 85660 }, { "grad_norm": 0.19733870029449463, "learning_rate": 5.510636969011407e-06, "loss": 0.0083, "step": 85670 }, { "grad_norm": 0.2553935647010803, "learning_rate": 5.503093370033096e-06, "loss": 0.0138, "step": 85680 }, { "grad_norm": 0.16754628717899323, "learning_rate": 5.495554637169603e-06, "loss": 0.0093, "step": 85690 }, { "grad_norm": 0.12378387153148651, "learning_rate": 5.488020771245328e-06, "loss": 0.0109, "step": 85700 }, { "grad_norm": 0.08114269375801086, "learning_rate": 5.480491773084195e-06, "loss": 0.0099, "step": 85710 }, { "grad_norm": 0.10204386711120605, "learning_rate": 5.472967643509525e-06, "loss": 0.009, "step": 85720 }, { "grad_norm": 0.07756543904542923, "learning_rate": 5.46544838334419e-06, "loss": 0.0111, "step": 85730 }, { "grad_norm": 0.1258767545223236, "learning_rate": 5.4579339934104525e-06, "loss": 0.0133, "step": 85740 }, { "grad_norm": 0.24544481933116913, "learning_rate": 5.450424474530091e-06, "loss": 0.0125, "step": 85750 }, { "grad_norm": 0.1633208990097046, "learning_rate": 5.442919827524323e-06, "loss": 0.0064, "step": 85760 }, { "grad_norm": 0.3066626191139221, "learning_rate": 5.435420053213863e-06, "loss": 0.0099, "step": 85770 }, { "grad_norm": 0.2884301245212555, "learning_rate": 5.427925152418845e-06, "loss": 0.0108, "step": 85780 }, { "grad_norm": 0.12704958021640778, "learning_rate": 5.420435125958934e-06, "loss": 0.0129, "step": 85790 }, { "grad_norm": 0.13777108490467072, "learning_rate": 5.412949974653209e-06, "loss": 0.0135, "step": 85800 }, { "grad_norm": 0.059419870376586914, "learning_rate": 5.405469699320248e-06, "loss": 0.0079, "step": 85810 }, { "grad_norm": 0.13969063758850098, "learning_rate": 5.397994300778064e-06, "loss": 0.0115, "step": 85820 }, { "grad_norm": 0.1724914312362671, "learning_rate": 5.39052377984417e-06, "loss": 0.0088, "step": 85830 }, { "grad_norm": 0.2846897542476654, "learning_rate": 5.383058137335523e-06, "loss": 0.012, "step": 85840 }, { "grad_norm": 0.0938645750284195, "learning_rate": 5.375597374068553e-06, "loss": 0.017, "step": 85850 }, { "grad_norm": 0.2190074920654297, "learning_rate": 5.368141490859163e-06, "loss": 0.011, "step": 85860 }, { "grad_norm": 0.0902981385588646, "learning_rate": 5.360690488522729e-06, "loss": 0.0117, "step": 85870 }, { "grad_norm": 0.3214481472969055, "learning_rate": 5.353244367874055e-06, "loss": 0.0147, "step": 85880 }, { "grad_norm": 0.0644555315375328, "learning_rate": 5.345803129727456e-06, "loss": 0.0109, "step": 85890 }, { "grad_norm": 0.1608390063047409, "learning_rate": 5.338366774896697e-06, "loss": 0.007, "step": 85900 }, { "grad_norm": 0.19830423593521118, "learning_rate": 5.330935304194995e-06, "loss": 0.0141, "step": 85910 }, { "grad_norm": 0.122417151927948, "learning_rate": 5.323508718435044e-06, "loss": 0.0083, "step": 85920 }, { "grad_norm": 0.2323331981897354, "learning_rate": 5.316087018429011e-06, "loss": 0.0119, "step": 85930 }, { "grad_norm": 0.14210627973079681, "learning_rate": 5.30867020498853e-06, "loss": 0.0131, "step": 85940 }, { "grad_norm": 0.14402304589748383, "learning_rate": 5.30125827892467e-06, "loss": 0.007, "step": 85950 }, { "grad_norm": 0.07389187067747116, "learning_rate": 5.293851241048015e-06, "loss": 0.0138, "step": 85960 }, { "grad_norm": 0.21655091643333435, "learning_rate": 5.286449092168555e-06, "loss": 0.0174, "step": 85970 }, { "grad_norm": 0.39409807324409485, "learning_rate": 5.2790518330958105e-06, "loss": 0.0087, "step": 85980 }, { "grad_norm": 0.13977374136447906, "learning_rate": 5.271659464638717e-06, "loss": 0.0119, "step": 85990 }, { "grad_norm": 0.2324511855840683, "learning_rate": 5.264271987605701e-06, "loss": 0.0082, "step": 86000 }, { "grad_norm": 0.5274220705032349, "learning_rate": 5.256889402804632e-06, "loss": 0.0177, "step": 86010 }, { "grad_norm": 0.0888272374868393, "learning_rate": 5.249511711042881e-06, "loss": 0.0082, "step": 86020 }, { "grad_norm": 0.4299010932445526, "learning_rate": 5.242138913127237e-06, "loss": 0.0163, "step": 86030 }, { "grad_norm": 0.2616226077079773, "learning_rate": 5.234771009863987e-06, "loss": 0.0112, "step": 86040 }, { "grad_norm": 0.14641353487968445, "learning_rate": 5.227408002058881e-06, "loss": 0.02, "step": 86050 }, { "grad_norm": 0.18784652650356293, "learning_rate": 5.220049890517126e-06, "loss": 0.0145, "step": 86060 }, { "grad_norm": 0.08683483302593231, "learning_rate": 5.212696676043377e-06, "loss": 0.0171, "step": 86070 }, { "grad_norm": 0.42730581760406494, "learning_rate": 5.205348359441798e-06, "loss": 0.0097, "step": 86080 }, { "grad_norm": 0.23542442917823792, "learning_rate": 5.198004941515966e-06, "loss": 0.0092, "step": 86090 }, { "grad_norm": 0.09416307508945465, "learning_rate": 5.190666423068952e-06, "loss": 0.0052, "step": 86100 }, { "grad_norm": 0.15160325169563293, "learning_rate": 5.183332804903284e-06, "loss": 0.0072, "step": 86110 }, { "grad_norm": 0.19648098945617676, "learning_rate": 5.176004087820979e-06, "loss": 0.0094, "step": 86120 }, { "grad_norm": 0.1292906552553177, "learning_rate": 5.168680272623455e-06, "loss": 0.0131, "step": 86130 }, { "grad_norm": 0.06549887359142303, "learning_rate": 5.161361360111661e-06, "loss": 0.0176, "step": 86140 }, { "grad_norm": 0.14629483222961426, "learning_rate": 5.154047351085984e-06, "loss": 0.0093, "step": 86150 }, { "grad_norm": 0.07460153847932816, "learning_rate": 5.1467382463462385e-06, "loss": 0.0062, "step": 86160 }, { "grad_norm": 0.04873330891132355, "learning_rate": 5.139434046691788e-06, "loss": 0.0073, "step": 86170 }, { "grad_norm": 0.0616617277264595, "learning_rate": 5.13213475292137e-06, "loss": 0.0095, "step": 86180 }, { "grad_norm": 0.5179033875465393, "learning_rate": 5.12484036583325e-06, "loss": 0.0151, "step": 86190 }, { "grad_norm": 0.20476004481315613, "learning_rate": 5.117550886225109e-06, "loss": 0.0118, "step": 86200 }, { "grad_norm": 0.04739360138773918, "learning_rate": 5.11026631489413e-06, "loss": 0.0056, "step": 86210 }, { "grad_norm": 0.4611720144748688, "learning_rate": 5.102986652636926e-06, "loss": 0.0148, "step": 86220 }, { "grad_norm": 0.2212892472743988, "learning_rate": 5.095711900249606e-06, "loss": 0.0094, "step": 86230 }, { "grad_norm": 0.14588099718093872, "learning_rate": 5.088442058527715e-06, "loss": 0.0128, "step": 86240 }, { "grad_norm": 0.28407081961631775, "learning_rate": 5.08117712826629e-06, "loss": 0.0138, "step": 86250 }, { "grad_norm": 0.12374317646026611, "learning_rate": 5.073917110259785e-06, "loss": 0.0147, "step": 86260 }, { "grad_norm": 0.06788095086812973, "learning_rate": 5.066662005302175e-06, "loss": 0.0084, "step": 86270 }, { "grad_norm": 0.021408846601843834, "learning_rate": 5.059411814186838e-06, "loss": 0.0089, "step": 86280 }, { "grad_norm": 0.08957773447036743, "learning_rate": 5.05216653770666e-06, "loss": 0.0188, "step": 86290 }, { "grad_norm": 0.20057237148284912, "learning_rate": 5.044926176653969e-06, "loss": 0.0109, "step": 86300 }, { "grad_norm": 0.23738843202590942, "learning_rate": 5.037690731820571e-06, "loss": 0.0083, "step": 86310 }, { "grad_norm": 0.14106497168540955, "learning_rate": 5.030460203997711e-06, "loss": 0.0097, "step": 86320 }, { "grad_norm": 0.22088976204395294, "learning_rate": 5.0232345939761174e-06, "loss": 0.0147, "step": 86330 }, { "grad_norm": 0.09952894598245621, "learning_rate": 5.016013902545957e-06, "loss": 0.0171, "step": 86340 }, { "grad_norm": 0.04386921226978302, "learning_rate": 5.008798130496889e-06, "loss": 0.0101, "step": 86350 }, { "grad_norm": 0.10810146480798721, "learning_rate": 5.001587278618008e-06, "loss": 0.0072, "step": 86360 }, { "grad_norm": 0.2564513087272644, "learning_rate": 4.994381347697891e-06, "loss": 0.0142, "step": 86370 }, { "grad_norm": 0.040322769433259964, "learning_rate": 4.987180338524572e-06, "loss": 0.0088, "step": 86380 }, { "grad_norm": 0.0589970238506794, "learning_rate": 4.9799842518855275e-06, "loss": 0.0059, "step": 86390 }, { "grad_norm": 0.0937858372926712, "learning_rate": 4.9727930885677285e-06, "loss": 0.0096, "step": 86400 }, { "grad_norm": 0.12851835787296295, "learning_rate": 4.965606849357573e-06, "loss": 0.0114, "step": 86410 }, { "grad_norm": 0.23647066950798035, "learning_rate": 4.9584255350409445e-06, "loss": 0.0089, "step": 86420 }, { "grad_norm": 0.25465700030326843, "learning_rate": 4.951249146403175e-06, "loss": 0.0169, "step": 86430 }, { "grad_norm": 0.07464119791984558, "learning_rate": 4.9440776842290815e-06, "loss": 0.0134, "step": 86440 }, { "grad_norm": 0.052372343838214874, "learning_rate": 4.936911149302903e-06, "loss": 0.0159, "step": 86450 }, { "grad_norm": 0.0333930179476738, "learning_rate": 4.929749542408379e-06, "loss": 0.0117, "step": 86460 }, { "grad_norm": 0.2220194935798645, "learning_rate": 4.922592864328673e-06, "loss": 0.008, "step": 86470 }, { "grad_norm": 0.0611792728304863, "learning_rate": 4.915441115846436e-06, "loss": 0.0083, "step": 86480 }, { "grad_norm": 0.20753179490566254, "learning_rate": 4.908294297743782e-06, "loss": 0.0219, "step": 86490 }, { "grad_norm": 0.16847367584705353, "learning_rate": 4.901152410802273e-06, "loss": 0.0113, "step": 86500 }, { "grad_norm": 0.030941203236579895, "learning_rate": 4.894015455802919e-06, "loss": 0.0128, "step": 86510 }, { "grad_norm": 0.37673959136009216, "learning_rate": 4.8868834335262346e-06, "loss": 0.0122, "step": 86520 }, { "grad_norm": 0.11731969565153122, "learning_rate": 4.87975634475214e-06, "loss": 0.0107, "step": 86530 }, { "grad_norm": 0.08065646141767502, "learning_rate": 4.8726341902600614e-06, "loss": 0.0143, "step": 86540 }, { "grad_norm": 0.09590427577495575, "learning_rate": 4.865516970828843e-06, "loss": 0.0103, "step": 86550 }, { "grad_norm": 0.18192288279533386, "learning_rate": 4.85840468723685e-06, "loss": 0.0108, "step": 86560 }, { "grad_norm": 0.3428710103034973, "learning_rate": 4.85129734026184e-06, "loss": 0.0126, "step": 86570 }, { "grad_norm": 0.3331150710582733, "learning_rate": 4.844194930681073e-06, "loss": 0.0133, "step": 86580 }, { "grad_norm": 0.14389026165008545, "learning_rate": 4.837097459271267e-06, "loss": 0.0134, "step": 86590 }, { "grad_norm": 0.12527503073215485, "learning_rate": 4.830004926808573e-06, "loss": 0.0132, "step": 86600 }, { "grad_norm": 0.09084448218345642, "learning_rate": 4.822917334068627e-06, "loss": 0.0108, "step": 86610 }, { "grad_norm": 0.4558441638946533, "learning_rate": 4.81583468182652e-06, "loss": 0.0067, "step": 86620 }, { "grad_norm": 0.21960890293121338, "learning_rate": 4.808756970856809e-06, "loss": 0.0114, "step": 86630 }, { "grad_norm": 0.10940748453140259, "learning_rate": 4.801684201933476e-06, "loss": 0.0095, "step": 86640 }, { "grad_norm": 0.16153286397457123, "learning_rate": 4.794616375830019e-06, "loss": 0.0078, "step": 86650 }, { "grad_norm": 0.42621785402297974, "learning_rate": 4.787553493319341e-06, "loss": 0.0105, "step": 86660 }, { "grad_norm": 0.27842456102371216, "learning_rate": 4.780495555173836e-06, "loss": 0.0142, "step": 86670 }, { "grad_norm": 0.14059577882289886, "learning_rate": 4.7734425621653475e-06, "loss": 0.0134, "step": 86680 }, { "grad_norm": 0.3678906559944153, "learning_rate": 4.766394515065198e-06, "loss": 0.0129, "step": 86690 }, { "grad_norm": 0.12489957362413406, "learning_rate": 4.75935141464412e-06, "loss": 0.0081, "step": 86700 }, { "grad_norm": 0.11016660183668137, "learning_rate": 4.75231326167237e-06, "loss": 0.0078, "step": 86710 }, { "grad_norm": 0.13337647914886475, "learning_rate": 4.745280056919599e-06, "loss": 0.0117, "step": 86720 }, { "grad_norm": 0.08359651267528534, "learning_rate": 4.738251801154975e-06, "loss": 0.0169, "step": 86730 }, { "grad_norm": 0.2982594668865204, "learning_rate": 4.731228495147061e-06, "loss": 0.0114, "step": 86740 }, { "grad_norm": 0.1748175472021103, "learning_rate": 4.7242101396639595e-06, "loss": 0.01, "step": 86750 }, { "grad_norm": 0.10176818817853928, "learning_rate": 4.717196735473151e-06, "loss": 0.009, "step": 86760 }, { "grad_norm": 0.1275169551372528, "learning_rate": 4.710188283341643e-06, "loss": 0.0093, "step": 86770 }, { "grad_norm": 0.14641323685646057, "learning_rate": 4.703184784035835e-06, "loss": 0.0178, "step": 86780 }, { "grad_norm": 0.06666991859674454, "learning_rate": 4.696186238321642e-06, "loss": 0.0076, "step": 86790 }, { "grad_norm": 0.24306493997573853, "learning_rate": 4.689192646964408e-06, "loss": 0.0155, "step": 86800 }, { "grad_norm": 0.26732775568962097, "learning_rate": 4.682204010728952e-06, "loss": 0.0095, "step": 86810 }, { "grad_norm": 0.17467831075191498, "learning_rate": 4.67522033037952e-06, "loss": 0.0111, "step": 86820 }, { "grad_norm": 0.08703286200761795, "learning_rate": 4.668241606679846e-06, "loss": 0.0106, "step": 86830 }, { "grad_norm": 0.2812419533729553, "learning_rate": 4.66126784039313e-06, "loss": 0.0151, "step": 86840 }, { "grad_norm": 0.25426173210144043, "learning_rate": 4.6542990322819836e-06, "loss": 0.0107, "step": 86850 }, { "grad_norm": 0.06877832114696503, "learning_rate": 4.64733518310852e-06, "loss": 0.0122, "step": 86860 }, { "grad_norm": 0.13254410028457642, "learning_rate": 4.640376293634291e-06, "loss": 0.0109, "step": 86870 }, { "grad_norm": 0.09965033829212189, "learning_rate": 4.633422364620327e-06, "loss": 0.0131, "step": 86880 }, { "grad_norm": 0.08661513775587082, "learning_rate": 4.6264733968270745e-06, "loss": 0.0082, "step": 86890 }, { "grad_norm": 0.1564328372478485, "learning_rate": 4.619529391014482e-06, "loss": 0.0199, "step": 86900 }, { "grad_norm": 0.24060553312301636, "learning_rate": 4.61259034794192e-06, "loss": 0.0124, "step": 86910 }, { "grad_norm": 0.37597838044166565, "learning_rate": 4.605656268368247e-06, "loss": 0.0114, "step": 86920 }, { "grad_norm": 0.1005011722445488, "learning_rate": 4.598727153051735e-06, "loss": 0.0182, "step": 86930 }, { "grad_norm": 0.21973817050457, "learning_rate": 4.591803002750178e-06, "loss": 0.0107, "step": 86940 }, { "grad_norm": 0.1778295338153839, "learning_rate": 4.58488381822077e-06, "loss": 0.0094, "step": 86950 }, { "grad_norm": 0.5544000864028931, "learning_rate": 4.5779696002201946e-06, "loss": 0.0125, "step": 86960 }, { "grad_norm": 0.08577487617731094, "learning_rate": 4.571060349504563e-06, "loss": 0.0125, "step": 86970 }, { "grad_norm": 0.1008368581533432, "learning_rate": 4.564156066829478e-06, "loss": 0.0173, "step": 86980 }, { "grad_norm": 0.2142457515001297, "learning_rate": 4.557256752949957e-06, "loss": 0.0099, "step": 86990 }, { "grad_norm": 0.08448533713817596, "learning_rate": 4.550362408620534e-06, "loss": 0.0137, "step": 87000 }, { "grad_norm": 0.09572944790124893, "learning_rate": 4.543473034595136e-06, "loss": 0.015, "step": 87010 }, { "grad_norm": 0.23876294493675232, "learning_rate": 4.536588631627192e-06, "loss": 0.0091, "step": 87020 }, { "grad_norm": 0.10192190110683441, "learning_rate": 4.5297092004695565e-06, "loss": 0.0121, "step": 87030 }, { "grad_norm": 0.2079932987689972, "learning_rate": 4.522834741874555e-06, "loss": 0.0123, "step": 87040 }, { "grad_norm": 0.1937623918056488, "learning_rate": 4.515965256593974e-06, "loss": 0.0109, "step": 87050 }, { "grad_norm": 0.15042966604232788, "learning_rate": 4.509100745379058e-06, "loss": 0.0123, "step": 87060 }, { "grad_norm": 0.03793735057115555, "learning_rate": 4.502241208980479e-06, "loss": 0.0088, "step": 87070 }, { "grad_norm": 0.14802400767803192, "learning_rate": 4.495386648148398e-06, "loss": 0.008, "step": 87080 }, { "grad_norm": 0.10114516317844391, "learning_rate": 4.488537063632431e-06, "loss": 0.0117, "step": 87090 }, { "grad_norm": 0.18476659059524536, "learning_rate": 4.481692456181608e-06, "loss": 0.0081, "step": 87100 }, { "grad_norm": 0.17406316101551056, "learning_rate": 4.474852826544479e-06, "loss": 0.015, "step": 87110 }, { "grad_norm": 0.262651652097702, "learning_rate": 4.468018175468974e-06, "loss": 0.0201, "step": 87120 }, { "grad_norm": 0.09602995961904526, "learning_rate": 4.461188503702568e-06, "loss": 0.0097, "step": 87130 }, { "grad_norm": 0.26669394969940186, "learning_rate": 4.454363811992107e-06, "loss": 0.0077, "step": 87140 }, { "grad_norm": 0.0961761474609375, "learning_rate": 4.447544101083956e-06, "loss": 0.0132, "step": 87150 }, { "grad_norm": 0.27692684531211853, "learning_rate": 4.440729371723884e-06, "loss": 0.0129, "step": 87160 }, { "grad_norm": 0.2579920291900635, "learning_rate": 4.433919624657162e-06, "loss": 0.0098, "step": 87170 }, { "grad_norm": 0.10520289093255997, "learning_rate": 4.427114860628462e-06, "loss": 0.0109, "step": 87180 }, { "grad_norm": 0.18930503726005554, "learning_rate": 4.420315080381982e-06, "loss": 0.0145, "step": 87190 }, { "grad_norm": 0.18404348194599152, "learning_rate": 4.4135202846613045e-06, "loss": 0.0123, "step": 87200 }, { "grad_norm": 0.10289274901151657, "learning_rate": 4.406730474209525e-06, "loss": 0.0083, "step": 87210 }, { "grad_norm": 0.3744080364704132, "learning_rate": 4.399945649769138e-06, "loss": 0.013, "step": 87220 }, { "grad_norm": 0.20818491280078888, "learning_rate": 4.393165812082151e-06, "loss": 0.0134, "step": 87230 }, { "grad_norm": 0.3155149221420288, "learning_rate": 4.386390961889969e-06, "loss": 0.0254, "step": 87240 }, { "grad_norm": 0.17622137069702148, "learning_rate": 4.3796210999334895e-06, "loss": 0.0086, "step": 87250 }, { "grad_norm": 0.22814074158668518, "learning_rate": 4.372856226953065e-06, "loss": 0.0088, "step": 87260 }, { "grad_norm": 0.04028722643852234, "learning_rate": 4.366096343688486e-06, "loss": 0.011, "step": 87270 }, { "grad_norm": 0.19901657104492188, "learning_rate": 4.359341450878995e-06, "loss": 0.009, "step": 87280 }, { "grad_norm": 0.06834934651851654, "learning_rate": 4.352591549263302e-06, "loss": 0.0108, "step": 87290 }, { "grad_norm": 0.17957668006420135, "learning_rate": 4.345846639579582e-06, "loss": 0.013, "step": 87300 }, { "grad_norm": 0.17444509267807007, "learning_rate": 4.339106722565417e-06, "loss": 0.0135, "step": 87310 }, { "grad_norm": 0.27233490347862244, "learning_rate": 4.332371798957896e-06, "loss": 0.013, "step": 87320 }, { "grad_norm": 0.33179715275764465, "learning_rate": 4.3256418694935295e-06, "loss": 0.0152, "step": 87330 }, { "grad_norm": 0.12823623418807983, "learning_rate": 4.3189169349083125e-06, "loss": 0.0106, "step": 87340 }, { "grad_norm": 0.1502694934606552, "learning_rate": 4.312196995937645e-06, "loss": 0.0079, "step": 87350 }, { "grad_norm": 0.3150097131729126, "learning_rate": 4.305482053316434e-06, "loss": 0.0108, "step": 87360 }, { "grad_norm": 0.2246718406677246, "learning_rate": 4.2987721077789925e-06, "loss": 0.0126, "step": 87370 }, { "grad_norm": 0.2650645971298218, "learning_rate": 4.2920671600591325e-06, "loss": 0.0079, "step": 87380 }, { "grad_norm": 0.08528195321559906, "learning_rate": 4.285367210890084e-06, "loss": 0.0083, "step": 87390 }, { "grad_norm": 0.3009548485279083, "learning_rate": 4.278672261004551e-06, "loss": 0.0155, "step": 87400 }, { "grad_norm": 0.13084179162979126, "learning_rate": 4.2719823111346704e-06, "loss": 0.0105, "step": 87410 }, { "grad_norm": 0.3071565330028534, "learning_rate": 4.265297362012061e-06, "loss": 0.0133, "step": 87420 }, { "grad_norm": 0.1100836992263794, "learning_rate": 4.258617414367766e-06, "loss": 0.0139, "step": 87430 }, { "grad_norm": 0.09277310967445374, "learning_rate": 4.251942468932296e-06, "loss": 0.0134, "step": 87440 }, { "grad_norm": 0.15683762729167938, "learning_rate": 4.2452725264356175e-06, "loss": 0.0144, "step": 87450 }, { "grad_norm": 0.1559327244758606, "learning_rate": 4.2386075876071575e-06, "loss": 0.0087, "step": 87460 }, { "grad_norm": 0.48237520456314087, "learning_rate": 4.231947653175755e-06, "loss": 0.0086, "step": 87470 }, { "grad_norm": 0.09184141457080841, "learning_rate": 4.225292723869761e-06, "loss": 0.0129, "step": 87480 }, { "grad_norm": 0.20985889434814453, "learning_rate": 4.218642800416916e-06, "loss": 0.0156, "step": 87490 }, { "grad_norm": 0.10099700838327408, "learning_rate": 4.211997883544472e-06, "loss": 0.0089, "step": 87500 }, { "grad_norm": 0.1927030235528946, "learning_rate": 4.205357973979096e-06, "loss": 0.0124, "step": 87510 }, { "grad_norm": 0.2543396055698395, "learning_rate": 4.198723072446925e-06, "loss": 0.0088, "step": 87520 }, { "grad_norm": 0.2677112817764282, "learning_rate": 4.192093179673534e-06, "loss": 0.008, "step": 87530 }, { "grad_norm": 0.14067557454109192, "learning_rate": 4.185468296383959e-06, "loss": 0.011, "step": 87540 }, { "grad_norm": 0.26347848773002625, "learning_rate": 4.1788484233027005e-06, "loss": 0.0136, "step": 87550 }, { "grad_norm": 0.4427027702331543, "learning_rate": 4.172233561153677e-06, "loss": 0.0086, "step": 87560 }, { "grad_norm": 0.15812614560127258, "learning_rate": 4.165623710660288e-06, "loss": 0.0117, "step": 87570 }, { "grad_norm": 0.04655829817056656, "learning_rate": 4.1590188725453835e-06, "loss": 0.0081, "step": 87580 }, { "grad_norm": 0.0723537728190422, "learning_rate": 4.152419047531259e-06, "loss": 0.0138, "step": 87590 }, { "grad_norm": 0.22678999602794647, "learning_rate": 4.145824236339651e-06, "loss": 0.0129, "step": 87600 }, { "grad_norm": 0.23804286122322083, "learning_rate": 4.1392344396917695e-06, "loss": 0.0103, "step": 87610 }, { "grad_norm": 0.16352833807468414, "learning_rate": 4.132649658308252e-06, "loss": 0.0206, "step": 87620 }, { "grad_norm": 0.838542640209198, "learning_rate": 4.126069892909213e-06, "loss": 0.0103, "step": 87630 }, { "grad_norm": 0.07591253519058228, "learning_rate": 4.119495144214191e-06, "loss": 0.0133, "step": 87640 }, { "grad_norm": 0.16021230816841125, "learning_rate": 4.112925412942214e-06, "loss": 0.0155, "step": 87650 }, { "grad_norm": 0.051610130816698074, "learning_rate": 4.106360699811712e-06, "loss": 0.0109, "step": 87660 }, { "grad_norm": 0.0372801348567009, "learning_rate": 4.099801005540616e-06, "loss": 0.0127, "step": 87670 }, { "grad_norm": 0.18271194398403168, "learning_rate": 4.093246330846256e-06, "loss": 0.0141, "step": 87680 }, { "grad_norm": 0.10421673953533173, "learning_rate": 4.086696676445467e-06, "loss": 0.014, "step": 87690 }, { "grad_norm": 0.11682233959436417, "learning_rate": 4.0801520430544914e-06, "loss": 0.0131, "step": 87700 }, { "grad_norm": 0.09486714750528336, "learning_rate": 4.073612431389062e-06, "loss": 0.014, "step": 87710 }, { "grad_norm": 0.10030561685562134, "learning_rate": 4.067077842164319e-06, "loss": 0.0169, "step": 87720 }, { "grad_norm": 0.10588516294956207, "learning_rate": 4.0605482760948945e-06, "loss": 0.0179, "step": 87730 }, { "grad_norm": 0.1671016961336136, "learning_rate": 4.054023733894829e-06, "loss": 0.0109, "step": 87740 }, { "grad_norm": 0.10982364416122437, "learning_rate": 4.047504216277653e-06, "loss": 0.012, "step": 87750 }, { "grad_norm": 0.5736903548240662, "learning_rate": 4.040989723956323e-06, "loss": 0.0165, "step": 87760 }, { "grad_norm": 0.2017136663198471, "learning_rate": 4.03448025764327e-06, "loss": 0.0091, "step": 87770 }, { "grad_norm": 0.15782390534877777, "learning_rate": 4.027975818050339e-06, "loss": 0.0115, "step": 87780 }, { "grad_norm": 0.17939956486225128, "learning_rate": 4.021476405888858e-06, "loss": 0.0193, "step": 87790 }, { "grad_norm": 0.37215644121170044, "learning_rate": 4.01498202186959e-06, "loss": 0.0101, "step": 87800 }, { "grad_norm": 0.26393580436706543, "learning_rate": 4.008492666702751e-06, "loss": 0.0156, "step": 87810 }, { "grad_norm": 0.25949862599372864, "learning_rate": 4.002008341098001e-06, "loss": 0.0114, "step": 87820 }, { "grad_norm": 0.2407839447259903, "learning_rate": 3.995529045764462e-06, "loss": 0.0095, "step": 87830 }, { "grad_norm": 0.09537072479724884, "learning_rate": 3.989054781410712e-06, "loss": 0.008, "step": 87840 }, { "grad_norm": 0.06752213090658188, "learning_rate": 3.982585548744744e-06, "loss": 0.0055, "step": 87850 }, { "grad_norm": 0.30404767394065857, "learning_rate": 3.9761213484740435e-06, "loss": 0.0096, "step": 87860 }, { "grad_norm": 0.21083970367908478, "learning_rate": 3.9696621813055055e-06, "loss": 0.0091, "step": 87870 }, { "grad_norm": 0.13429929316043854, "learning_rate": 3.9632080479455044e-06, "loss": 0.009, "step": 87880 }, { "grad_norm": 0.08023624122142792, "learning_rate": 3.956758949099859e-06, "loss": 0.0113, "step": 87890 }, { "grad_norm": 0.09528075903654099, "learning_rate": 3.9503148854738385e-06, "loss": 0.0087, "step": 87900 }, { "grad_norm": 0.050149254500865936, "learning_rate": 3.943875857772139e-06, "loss": 0.0251, "step": 87910 }, { "grad_norm": 0.06572186201810837, "learning_rate": 3.9374418666989375e-06, "loss": 0.0137, "step": 87920 }, { "grad_norm": 0.07732739299535751, "learning_rate": 3.931012912957838e-06, "loss": 0.0114, "step": 87930 }, { "grad_norm": 0.3051956295967102, "learning_rate": 3.9245889972519045e-06, "loss": 0.0177, "step": 87940 }, { "grad_norm": 0.1794399619102478, "learning_rate": 3.9181701202836265e-06, "loss": 0.0091, "step": 87950 }, { "grad_norm": 0.09104929864406586, "learning_rate": 3.911756282755003e-06, "loss": 0.0131, "step": 87960 }, { "grad_norm": 0.10380586981773376, "learning_rate": 3.905347485367411e-06, "loss": 0.0087, "step": 87970 }, { "grad_norm": 0.27828869223594666, "learning_rate": 3.898943728821725e-06, "loss": 0.0151, "step": 87980 }, { "grad_norm": 0.15897925198078156, "learning_rate": 3.892545013818227e-06, "loss": 0.0107, "step": 87990 }, { "grad_norm": 0.11430765688419342, "learning_rate": 3.886151341056687e-06, "loss": 0.0116, "step": 88000 }, { "grad_norm": 0.04160941392183304, "learning_rate": 3.8797627112363066e-06, "loss": 0.0077, "step": 88010 }, { "grad_norm": 0.178782120347023, "learning_rate": 3.873379125055743e-06, "loss": 0.0126, "step": 88020 }, { "grad_norm": 0.11536722630262375, "learning_rate": 3.867000583213082e-06, "loss": 0.0137, "step": 88030 }, { "grad_norm": 0.19980113208293915, "learning_rate": 3.860627086405871e-06, "loss": 0.0141, "step": 88040 }, { "grad_norm": 0.35312193632125854, "learning_rate": 3.854258635331126e-06, "loss": 0.0125, "step": 88050 }, { "grad_norm": 0.29005593061447144, "learning_rate": 3.8478952306852665e-06, "loss": 0.0138, "step": 88060 }, { "grad_norm": 0.38922080397605896, "learning_rate": 3.841536873164198e-06, "loss": 0.0128, "step": 88070 }, { "grad_norm": 0.14564085006713867, "learning_rate": 3.8351835634632626e-06, "loss": 0.0111, "step": 88080 }, { "grad_norm": 0.44546839594841003, "learning_rate": 3.828835302277245e-06, "loss": 0.012, "step": 88090 }, { "grad_norm": 0.14562153816223145, "learning_rate": 3.822492090300378e-06, "loss": 0.0117, "step": 88100 }, { "grad_norm": 0.2593776285648346, "learning_rate": 3.8161539282263566e-06, "loss": 0.0132, "step": 88110 }, { "grad_norm": 0.04728887602686882, "learning_rate": 3.8098208167482995e-06, "loss": 0.0075, "step": 88120 }, { "grad_norm": 0.3683117628097534, "learning_rate": 3.803492756558796e-06, "loss": 0.0098, "step": 88130 }, { "grad_norm": 0.05961553379893303, "learning_rate": 3.797169748349855e-06, "loss": 0.0106, "step": 88140 }, { "grad_norm": 0.28920984268188477, "learning_rate": 3.790851792812977e-06, "loss": 0.0104, "step": 88150 }, { "grad_norm": 0.14380212128162384, "learning_rate": 3.784538890639072e-06, "loss": 0.01, "step": 88160 }, { "grad_norm": 0.14802776277065277, "learning_rate": 3.7782310425185153e-06, "loss": 0.0088, "step": 88170 }, { "grad_norm": 0.2648669481277466, "learning_rate": 3.77192824914111e-06, "loss": 0.0102, "step": 88180 }, { "grad_norm": 0.17240791022777557, "learning_rate": 3.7656305111961333e-06, "loss": 0.0134, "step": 88190 }, { "grad_norm": 0.06408999115228653, "learning_rate": 3.7593378293722782e-06, "loss": 0.0108, "step": 88200 }, { "grad_norm": 0.03506195917725563, "learning_rate": 3.7530502043577333e-06, "loss": 0.0103, "step": 88210 }, { "grad_norm": 0.1246001198887825, "learning_rate": 3.746767636840076e-06, "loss": 0.0163, "step": 88220 }, { "grad_norm": 0.11059602349996567, "learning_rate": 3.74049012750638e-06, "loss": 0.0128, "step": 88230 }, { "grad_norm": 0.3062751591205597, "learning_rate": 3.734217677043128e-06, "loss": 0.0088, "step": 88240 }, { "grad_norm": 0.12255104631185532, "learning_rate": 3.727950286136267e-06, "loss": 0.0071, "step": 88250 }, { "grad_norm": 0.2809675335884094, "learning_rate": 3.7216879554711925e-06, "loss": 0.0112, "step": 88260 }, { "grad_norm": 0.14599978923797607, "learning_rate": 3.715430685732757e-06, "loss": 0.0183, "step": 88270 }, { "grad_norm": 0.370880663394928, "learning_rate": 3.7091784776052243e-06, "loss": 0.0106, "step": 88280 }, { "grad_norm": 0.0991148054599762, "learning_rate": 3.7029313317723356e-06, "loss": 0.0109, "step": 88290 }, { "grad_norm": 0.1562110334634781, "learning_rate": 3.6966892489172787e-06, "loss": 0.0132, "step": 88300 }, { "grad_norm": 0.1159905269742012, "learning_rate": 3.6904522297226575e-06, "loss": 0.0089, "step": 88310 }, { "grad_norm": 0.341277539730072, "learning_rate": 3.6842202748705647e-06, "loss": 0.016, "step": 88320 }, { "grad_norm": 0.1927431970834732, "learning_rate": 3.677993385042494e-06, "loss": 0.0133, "step": 88330 }, { "grad_norm": 0.09557413309812546, "learning_rate": 3.67177156091944e-06, "loss": 0.0155, "step": 88340 }, { "grad_norm": 0.10824337601661682, "learning_rate": 3.6655548031817866e-06, "loss": 0.0128, "step": 88350 }, { "grad_norm": 0.20324817299842834, "learning_rate": 3.6593431125094e-06, "loss": 0.0134, "step": 88360 }, { "grad_norm": 0.03409780561923981, "learning_rate": 3.6531364895815766e-06, "loss": 0.0084, "step": 88370 }, { "grad_norm": 0.17327970266342163, "learning_rate": 3.6469349350770677e-06, "loss": 0.0145, "step": 88380 }, { "grad_norm": 0.13989359140396118, "learning_rate": 3.6407384496740525e-06, "loss": 0.015, "step": 88390 }, { "grad_norm": 0.39681047201156616, "learning_rate": 3.6345470340501954e-06, "loss": 0.0107, "step": 88400 }, { "grad_norm": 0.35514307022094727, "learning_rate": 3.6283606888825537e-06, "loss": 0.0151, "step": 88410 }, { "grad_norm": 0.4191915988922119, "learning_rate": 3.622179414847676e-06, "loss": 0.0128, "step": 88420 }, { "grad_norm": 0.17231793701648712, "learning_rate": 3.616003212621527e-06, "loss": 0.0174, "step": 88430 }, { "grad_norm": 0.20459973812103271, "learning_rate": 3.609832082879533e-06, "loss": 0.0118, "step": 88440 }, { "grad_norm": 0.08930548280477524, "learning_rate": 3.6036660262965485e-06, "loss": 0.0157, "step": 88450 }, { "grad_norm": 0.12224181741476059, "learning_rate": 3.597505043546895e-06, "loss": 0.0085, "step": 88460 }, { "grad_norm": 0.1495826691389084, "learning_rate": 3.5913491353043227e-06, "loss": 0.0076, "step": 88470 }, { "grad_norm": 0.16771283745765686, "learning_rate": 3.5851983022420375e-06, "loss": 0.0118, "step": 88480 }, { "grad_norm": 0.0823882594704628, "learning_rate": 3.5790525450326797e-06, "loss": 0.0126, "step": 88490 }, { "grad_norm": 0.29132747650146484, "learning_rate": 3.572911864348344e-06, "loss": 0.0142, "step": 88500 }, { "grad_norm": 0.12544074654579163, "learning_rate": 3.5667762608605717e-06, "loss": 0.0066, "step": 88510 }, { "grad_norm": 0.15741120278835297, "learning_rate": 3.5606457352403255e-06, "loss": 0.0077, "step": 88520 }, { "grad_norm": 0.07929055392742157, "learning_rate": 3.554520288158042e-06, "loss": 0.0058, "step": 88530 }, { "grad_norm": 0.13865649700164795, "learning_rate": 3.5483999202835903e-06, "loss": 0.0084, "step": 88540 }, { "grad_norm": 0.031600482761859894, "learning_rate": 3.5422846322862913e-06, "loss": 0.0153, "step": 88550 }, { "grad_norm": 0.06421271711587906, "learning_rate": 3.5361744248348925e-06, "loss": 0.0077, "step": 88560 }, { "grad_norm": 0.26195284724235535, "learning_rate": 3.5300692985976047e-06, "loss": 0.0169, "step": 88570 }, { "grad_norm": 0.1765604317188263, "learning_rate": 3.52396925424206e-06, "loss": 0.0099, "step": 88580 }, { "grad_norm": 0.09686672687530518, "learning_rate": 3.5178742924353747e-06, "loss": 0.0084, "step": 88590 }, { "grad_norm": 0.11295813322067261, "learning_rate": 3.511784413844066e-06, "loss": 0.0127, "step": 88600 }, { "grad_norm": 0.06975176185369492, "learning_rate": 3.5056996191341287e-06, "loss": 0.0099, "step": 88610 }, { "grad_norm": 0.12964215874671936, "learning_rate": 3.4996199089709692e-06, "loss": 0.0102, "step": 88620 }, { "grad_norm": 0.2574251592159271, "learning_rate": 3.4935452840194725e-06, "loss": 0.0113, "step": 88630 }, { "grad_norm": 0.22698867321014404, "learning_rate": 3.48747574494393e-06, "loss": 0.0137, "step": 88640 }, { "grad_norm": 0.25553879141807556, "learning_rate": 3.4814112924081154e-06, "loss": 0.0127, "step": 88650 }, { "grad_norm": 0.19922605156898499, "learning_rate": 3.475351927075221e-06, "loss": 0.0082, "step": 88660 }, { "grad_norm": 0.07112530618906021, "learning_rate": 3.469297649607894e-06, "loss": 0.0119, "step": 88670 }, { "grad_norm": 0.16905662417411804, "learning_rate": 3.463248460668217e-06, "loss": 0.0077, "step": 88680 }, { "grad_norm": 0.239246666431427, "learning_rate": 3.457204360917726e-06, "loss": 0.0119, "step": 88690 }, { "grad_norm": 0.42335259914398193, "learning_rate": 3.4511653510173827e-06, "loss": 0.0103, "step": 88700 }, { "grad_norm": 0.19476988911628723, "learning_rate": 3.4451314316276074e-06, "loss": 0.0108, "step": 88710 }, { "grad_norm": 0.16517390310764313, "learning_rate": 3.4391026034082674e-06, "loss": 0.0099, "step": 88720 }, { "grad_norm": 0.2893005907535553, "learning_rate": 3.433078867018674e-06, "loss": 0.0072, "step": 88730 }, { "grad_norm": 0.37026745080947876, "learning_rate": 3.427060223117551e-06, "loss": 0.0108, "step": 88740 }, { "grad_norm": 0.08348280191421509, "learning_rate": 3.4210466723631042e-06, "loss": 0.0085, "step": 88750 }, { "grad_norm": 0.15422983467578888, "learning_rate": 3.415038215412969e-06, "loss": 0.011, "step": 88760 }, { "grad_norm": 0.08041785657405853, "learning_rate": 3.4090348529242088e-06, "loss": 0.0105, "step": 88770 }, { "grad_norm": 0.12859782576560974, "learning_rate": 3.4030365855533487e-06, "loss": 0.0106, "step": 88780 }, { "grad_norm": 0.18188640475273132, "learning_rate": 3.397043413956347e-06, "loss": 0.0106, "step": 88790 }, { "grad_norm": 0.07547745853662491, "learning_rate": 3.3910553387886237e-06, "loss": 0.0071, "step": 88800 }, { "grad_norm": 0.03346533328294754, "learning_rate": 3.3850723607049996e-06, "loss": 0.0078, "step": 88810 }, { "grad_norm": 0.07579745352268219, "learning_rate": 3.3790944803597902e-06, "loss": 0.0096, "step": 88820 }, { "grad_norm": 0.37150269746780396, "learning_rate": 3.373121698406706e-06, "loss": 0.0142, "step": 88830 }, { "grad_norm": 0.11529012769460678, "learning_rate": 3.3671540154989302e-06, "loss": 0.0093, "step": 88840 }, { "grad_norm": 0.09956041723489761, "learning_rate": 3.361191432289079e-06, "loss": 0.0121, "step": 88850 }, { "grad_norm": 0.06727882474660873, "learning_rate": 3.3552339494292207e-06, "loss": 0.0105, "step": 88860 }, { "grad_norm": 0.1926492601633072, "learning_rate": 3.349281567570839e-06, "loss": 0.0147, "step": 88870 }, { "grad_norm": 0.06580086797475815, "learning_rate": 3.3433342873648965e-06, "loss": 0.0135, "step": 88880 }, { "grad_norm": 0.18750101327896118, "learning_rate": 3.337392109461762e-06, "loss": 0.0091, "step": 88890 }, { "grad_norm": 0.10306583344936371, "learning_rate": 3.331455034511266e-06, "loss": 0.0143, "step": 88900 }, { "grad_norm": 0.034797247499227524, "learning_rate": 3.325523063162689e-06, "loss": 0.0075, "step": 88910 }, { "grad_norm": 0.4130328595638275, "learning_rate": 3.3195961960647393e-06, "loss": 0.015, "step": 88920 }, { "grad_norm": 0.20495477318763733, "learning_rate": 3.31367443386556e-06, "loss": 0.0176, "step": 88930 }, { "grad_norm": 0.1482263058423996, "learning_rate": 3.3077577772127556e-06, "loss": 0.0071, "step": 88940 }, { "grad_norm": 0.31759750843048096, "learning_rate": 3.301846226753358e-06, "loss": 0.0128, "step": 88950 }, { "grad_norm": 0.6265775561332703, "learning_rate": 3.2959397831338444e-06, "loss": 0.0105, "step": 88960 }, { "grad_norm": 0.41080179810523987, "learning_rate": 3.2900384470001432e-06, "loss": 0.0175, "step": 88970 }, { "grad_norm": 0.05322590097784996, "learning_rate": 3.28414221899761e-06, "loss": 0.0106, "step": 88980 }, { "grad_norm": 0.2579510807991028, "learning_rate": 3.278251099771046e-06, "loss": 0.0089, "step": 88990 }, { "grad_norm": 0.12620531022548676, "learning_rate": 3.2723650899646906e-06, "loss": 0.0246, "step": 89000 }, { "grad_norm": 0.10075929015874863, "learning_rate": 3.2664841902222466e-06, "loss": 0.0056, "step": 89010 }, { "grad_norm": 0.46553459763526917, "learning_rate": 3.2606084011868208e-06, "loss": 0.0123, "step": 89020 }, { "grad_norm": 0.27228474617004395, "learning_rate": 3.254737723500989e-06, "loss": 0.012, "step": 89030 }, { "grad_norm": 0.2489703893661499, "learning_rate": 3.2488721578067595e-06, "loss": 0.01, "step": 89040 }, { "grad_norm": 0.21269279718399048, "learning_rate": 3.243011704745591e-06, "loss": 0.0149, "step": 89050 }, { "grad_norm": 0.13551561534404755, "learning_rate": 3.237156364958355e-06, "loss": 0.0122, "step": 89060 }, { "grad_norm": 0.09772589802742004, "learning_rate": 3.231306139085394e-06, "loss": 0.0107, "step": 89070 }, { "grad_norm": 0.11602994054555893, "learning_rate": 3.2254610277664797e-06, "loss": 0.007, "step": 89080 }, { "grad_norm": 0.11921004951000214, "learning_rate": 3.2196210316408183e-06, "loss": 0.0144, "step": 89090 }, { "grad_norm": 0.14969374239444733, "learning_rate": 3.2137861513470644e-06, "loss": 0.0085, "step": 89100 }, { "grad_norm": 0.1418568342924118, "learning_rate": 3.2079563875233252e-06, "loss": 0.0083, "step": 89110 }, { "grad_norm": 0.07643872499465942, "learning_rate": 3.2021317408071184e-06, "loss": 0.0092, "step": 89120 }, { "grad_norm": 0.15410654246807098, "learning_rate": 3.196312211835434e-06, "loss": 0.0123, "step": 89130 }, { "grad_norm": 0.2428254336118698, "learning_rate": 3.1904978012446686e-06, "loss": 0.0144, "step": 89140 }, { "grad_norm": 0.3814246952533722, "learning_rate": 3.184688509670691e-06, "loss": 0.0113, "step": 89150 }, { "grad_norm": 0.12330566346645355, "learning_rate": 3.178884337748783e-06, "loss": 0.0094, "step": 89160 }, { "grad_norm": 0.2006119042634964, "learning_rate": 3.173085286113697e-06, "loss": 0.0088, "step": 89170 }, { "grad_norm": 0.09115991741418839, "learning_rate": 3.1672913553995986e-06, "loss": 0.008, "step": 89180 }, { "grad_norm": 0.06357196718454361, "learning_rate": 3.1615025462401138e-06, "loss": 0.0102, "step": 89190 }, { "grad_norm": 0.33022019267082214, "learning_rate": 3.1557188592682808e-06, "loss": 0.009, "step": 89200 }, { "grad_norm": 0.21158531308174133, "learning_rate": 3.1499402951166045e-06, "loss": 0.0092, "step": 89210 }, { "grad_norm": 0.14186806976795197, "learning_rate": 3.144166854417019e-06, "loss": 0.0134, "step": 89220 }, { "grad_norm": 0.1065024584531784, "learning_rate": 3.1383985378009073e-06, "loss": 0.0131, "step": 89230 }, { "grad_norm": 0.2690989077091217, "learning_rate": 3.1326353458990654e-06, "loss": 0.0107, "step": 89240 }, { "grad_norm": 0.14513620734214783, "learning_rate": 3.126877279341761e-06, "loss": 0.0101, "step": 89250 }, { "grad_norm": 0.16976945102214813, "learning_rate": 3.1211243387586964e-06, "loss": 0.0143, "step": 89260 }, { "grad_norm": 0.1948350965976715, "learning_rate": 3.115376524778979e-06, "loss": 0.0084, "step": 89270 }, { "grad_norm": 0.3415849804878235, "learning_rate": 3.109633838031195e-06, "loss": 0.021, "step": 89280 }, { "grad_norm": 0.03734444081783295, "learning_rate": 3.103896279143359e-06, "loss": 0.012, "step": 89290 }, { "grad_norm": 0.33255547285079956, "learning_rate": 3.098163848742924e-06, "loss": 0.015, "step": 89300 }, { "grad_norm": 0.10384119302034378, "learning_rate": 3.0924365474567675e-06, "loss": 0.0098, "step": 89310 }, { "grad_norm": 0.20378264784812927, "learning_rate": 3.086714375911237e-06, "loss": 0.0114, "step": 89320 }, { "grad_norm": 0.12702180445194244, "learning_rate": 3.0809973347320775e-06, "loss": 0.0127, "step": 89330 }, { "grad_norm": 0.09932231158018112, "learning_rate": 3.075285424544516e-06, "loss": 0.0148, "step": 89340 }, { "grad_norm": 0.2645798623561859, "learning_rate": 3.069578645973176e-06, "loss": 0.0066, "step": 89350 }, { "grad_norm": 0.136709526181221, "learning_rate": 3.0638769996421735e-06, "loss": 0.0088, "step": 89360 }, { "grad_norm": 0.5113009810447693, "learning_rate": 3.058180486175005e-06, "loss": 0.0119, "step": 89370 }, { "grad_norm": 0.2501777112483978, "learning_rate": 3.052489106194645e-06, "loss": 0.0112, "step": 89380 }, { "grad_norm": 0.1446017175912857, "learning_rate": 3.0468028603234946e-06, "loss": 0.0086, "step": 89390 }, { "grad_norm": 0.10282614082098007, "learning_rate": 3.04112174918339e-06, "loss": 0.0086, "step": 89400 }, { "grad_norm": 0.3042159676551819, "learning_rate": 3.0354457733955954e-06, "loss": 0.0107, "step": 89410 }, { "grad_norm": 0.1813351958990097, "learning_rate": 3.029774933580859e-06, "loss": 0.0097, "step": 89420 }, { "grad_norm": 0.12590432167053223, "learning_rate": 3.0241092303593065e-06, "loss": 0.0169, "step": 89430 }, { "grad_norm": 0.271434485912323, "learning_rate": 3.018448664350554e-06, "loss": 0.0144, "step": 89440 }, { "grad_norm": 0.15984641015529633, "learning_rate": 3.0127932361736055e-06, "loss": 0.0117, "step": 89450 }, { "grad_norm": 0.25633931159973145, "learning_rate": 3.00714294644695e-06, "loss": 0.0158, "step": 89460 }, { "grad_norm": 0.3706885576248169, "learning_rate": 3.0014977957884828e-06, "loss": 0.0185, "step": 89470 }, { "grad_norm": 0.10001124441623688, "learning_rate": 2.99585778481557e-06, "loss": 0.0137, "step": 89480 }, { "grad_norm": 0.09909133613109589, "learning_rate": 2.990222914144969e-06, "loss": 0.0204, "step": 89490 }, { "grad_norm": 0.10137205570936203, "learning_rate": 2.984593184392914e-06, "loss": 0.0152, "step": 89500 }, { "grad_norm": 0.32730573415756226, "learning_rate": 2.978968596175069e-06, "loss": 0.0156, "step": 89510 }, { "grad_norm": 0.20274725556373596, "learning_rate": 2.9733491501065137e-06, "loss": 0.0083, "step": 89520 }, { "grad_norm": 0.24725507199764252, "learning_rate": 2.9677348468018005e-06, "loss": 0.0075, "step": 89530 }, { "grad_norm": 0.11677354574203491, "learning_rate": 2.9621256868748726e-06, "loss": 0.0081, "step": 89540 }, { "grad_norm": 0.16588860750198364, "learning_rate": 2.956521670939183e-06, "loss": 0.0081, "step": 89550 }, { "grad_norm": 0.12101361155509949, "learning_rate": 2.950922799607542e-06, "loss": 0.0197, "step": 89560 }, { "grad_norm": 0.7846910357475281, "learning_rate": 2.9453290734922535e-06, "loss": 0.0161, "step": 89570 }, { "grad_norm": 0.21583089232444763, "learning_rate": 2.939740493205029e-06, "loss": 0.0132, "step": 89580 }, { "grad_norm": 0.05484616383910179, "learning_rate": 2.9341570593570345e-06, "loss": 0.014, "step": 89590 }, { "grad_norm": 0.2588973939418793, "learning_rate": 2.928578772558843e-06, "loss": 0.014, "step": 89600 }, { "grad_norm": 0.05668545141816139, "learning_rate": 2.923005633420528e-06, "loss": 0.0117, "step": 89610 }, { "grad_norm": 0.14250881969928741, "learning_rate": 2.9174376425515237e-06, "loss": 0.0067, "step": 89620 }, { "grad_norm": 0.21503221988677979, "learning_rate": 2.91187480056076e-06, "loss": 0.014, "step": 89630 }, { "grad_norm": 0.38565194606781006, "learning_rate": 2.9063171080565664e-06, "loss": 0.0117, "step": 89640 }, { "grad_norm": 0.10370967537164688, "learning_rate": 2.900764565646741e-06, "loss": 0.0108, "step": 89650 }, { "grad_norm": 0.23641782999038696, "learning_rate": 2.895217173938475e-06, "loss": 0.0103, "step": 89660 }, { "grad_norm": 0.1528881937265396, "learning_rate": 2.8896749335384445e-06, "loss": 0.0164, "step": 89670 }, { "grad_norm": 0.06499778479337692, "learning_rate": 2.884137845052731e-06, "loss": 0.0086, "step": 89680 }, { "grad_norm": 0.13862548768520355, "learning_rate": 2.8786059090868734e-06, "loss": 0.0092, "step": 89690 }, { "grad_norm": 0.1012389212846756, "learning_rate": 2.87307912624582e-06, "loss": 0.0132, "step": 89700 }, { "grad_norm": 0.38808056712150574, "learning_rate": 2.867557497133977e-06, "loss": 0.0142, "step": 89710 }, { "grad_norm": 0.10734420269727707, "learning_rate": 2.862041022355194e-06, "loss": 0.0131, "step": 89720 }, { "grad_norm": 0.06111549213528633, "learning_rate": 2.8565297025127226e-06, "loss": 0.0078, "step": 89730 }, { "grad_norm": 0.5049020051956177, "learning_rate": 2.8510235382092853e-06, "loss": 0.0098, "step": 89740 }, { "grad_norm": 0.15718962252140045, "learning_rate": 2.845522530047029e-06, "loss": 0.0079, "step": 89750 }, { "grad_norm": 0.07414738833904266, "learning_rate": 2.8400266786275387e-06, "loss": 0.0113, "step": 89760 }, { "grad_norm": 0.2542971968650818, "learning_rate": 2.834535984551817e-06, "loss": 0.0118, "step": 89770 }, { "grad_norm": 0.15990376472473145, "learning_rate": 2.8290504484203397e-06, "loss": 0.0214, "step": 89780 }, { "grad_norm": 0.15093766152858734, "learning_rate": 2.8235700708329705e-06, "loss": 0.0116, "step": 89790 }, { "grad_norm": 0.23996223509311676, "learning_rate": 2.818094852389058e-06, "loss": 0.0224, "step": 89800 }, { "grad_norm": 0.03839951008558273, "learning_rate": 2.8126247936873516e-06, "loss": 0.0099, "step": 89810 }, { "grad_norm": 0.09301851689815521, "learning_rate": 2.8071598953260614e-06, "loss": 0.011, "step": 89820 }, { "grad_norm": 0.17515625059604645, "learning_rate": 2.8017001579028035e-06, "loss": 0.0096, "step": 89830 }, { "grad_norm": 0.15999987721443176, "learning_rate": 2.7962455820146617e-06, "loss": 0.0108, "step": 89840 }, { "grad_norm": 0.14632299542427063, "learning_rate": 2.7907961682581253e-06, "loss": 0.0111, "step": 89850 }, { "grad_norm": 0.1298254430294037, "learning_rate": 2.7853519172291453e-06, "loss": 0.0079, "step": 89860 }, { "grad_norm": 0.32432985305786133, "learning_rate": 2.7799128295230904e-06, "loss": 0.013, "step": 89870 }, { "grad_norm": 0.0692007765173912, "learning_rate": 2.774478905734784e-06, "loss": 0.0102, "step": 89880 }, { "grad_norm": 0.12032416462898254, "learning_rate": 2.7690501464584516e-06, "loss": 0.0114, "step": 89890 }, { "grad_norm": 0.13789233565330505, "learning_rate": 2.763626552287796e-06, "loss": 0.0093, "step": 89900 }, { "grad_norm": 0.26598793268203735, "learning_rate": 2.7582081238159083e-06, "loss": 0.0127, "step": 89910 }, { "grad_norm": 0.12792401015758514, "learning_rate": 2.7527948616353548e-06, "loss": 0.0119, "step": 89920 }, { "grad_norm": 0.3219212293624878, "learning_rate": 2.7473867663381226e-06, "loss": 0.007, "step": 89930 }, { "grad_norm": 0.08078218996524811, "learning_rate": 2.7419838385156383e-06, "loss": 0.0119, "step": 89940 }, { "grad_norm": 0.508324384689331, "learning_rate": 2.7365860787587407e-06, "loss": 0.0165, "step": 89950 }, { "grad_norm": 0.2584209144115448, "learning_rate": 2.7311934876577295e-06, "loss": 0.0177, "step": 89960 }, { "grad_norm": 0.18661291897296906, "learning_rate": 2.725806065802339e-06, "loss": 0.0188, "step": 89970 }, { "grad_norm": 0.10471780598163605, "learning_rate": 2.720423813781714e-06, "loss": 0.0065, "step": 89980 }, { "grad_norm": 0.18425124883651733, "learning_rate": 2.715046732184462e-06, "loss": 0.0164, "step": 89990 }, { "grad_norm": 0.14234724640846252, "learning_rate": 2.7096748215986013e-06, "loss": 0.0077, "step": 90000 }, { "grad_norm": 0.1331988424062729, "learning_rate": 2.704308082611612e-06, "loss": 0.0101, "step": 90010 }, { "grad_norm": 0.2922443151473999, "learning_rate": 2.698946515810374e-06, "loss": 0.0075, "step": 90020 }, { "grad_norm": 0.18019963800907135, "learning_rate": 2.6935901217812367e-06, "loss": 0.0104, "step": 90030 }, { "grad_norm": 0.18459661304950714, "learning_rate": 2.688238901109952e-06, "loss": 0.0124, "step": 90040 }, { "grad_norm": 0.13014273345470428, "learning_rate": 2.68289285438173e-06, "loss": 0.0092, "step": 90050 }, { "grad_norm": 0.2090766876935959, "learning_rate": 2.6775519821812033e-06, "loss": 0.0138, "step": 90060 }, { "grad_norm": 0.1506453901529312, "learning_rate": 2.672216285092449e-06, "loss": 0.0123, "step": 90070 }, { "grad_norm": 0.09257199615240097, "learning_rate": 2.666885763698962e-06, "loss": 0.0072, "step": 90080 }, { "grad_norm": 0.09941481798887253, "learning_rate": 2.661560418583692e-06, "loss": 0.0145, "step": 90090 }, { "grad_norm": 0.13337306678295135, "learning_rate": 2.656240250328995e-06, "loss": 0.0111, "step": 90100 }, { "grad_norm": 0.13529913127422333, "learning_rate": 2.650925259516679e-06, "loss": 0.0136, "step": 90110 }, { "grad_norm": 0.1897752583026886, "learning_rate": 2.645615446727995e-06, "loss": 0.0119, "step": 90120 }, { "grad_norm": 0.16394774615764618, "learning_rate": 2.6403108125436115e-06, "loss": 0.009, "step": 90130 }, { "grad_norm": 0.16236500442028046, "learning_rate": 2.6350113575436264e-06, "loss": 0.0079, "step": 90140 }, { "grad_norm": 0.06491877138614655, "learning_rate": 2.629717082307598e-06, "loss": 0.0072, "step": 90150 }, { "grad_norm": 0.37463173270225525, "learning_rate": 2.6244279874144793e-06, "loss": 0.0098, "step": 90160 }, { "grad_norm": 0.18848003447055817, "learning_rate": 2.6191440734426862e-06, "loss": 0.0157, "step": 90170 }, { "grad_norm": 0.10542196035385132, "learning_rate": 2.613865340970062e-06, "loss": 0.0089, "step": 90180 }, { "grad_norm": 0.14872531592845917, "learning_rate": 2.608591790573889e-06, "loss": 0.0092, "step": 90190 }, { "grad_norm": 0.09666827321052551, "learning_rate": 2.603323422830861e-06, "loss": 0.0084, "step": 90200 }, { "grad_norm": 0.2057550549507141, "learning_rate": 2.5980602383171183e-06, "loss": 0.0114, "step": 90210 }, { "grad_norm": 0.14090117812156677, "learning_rate": 2.5928022376082496e-06, "loss": 0.0087, "step": 90220 }, { "grad_norm": 0.7017314434051514, "learning_rate": 2.5875494212792505e-06, "loss": 0.0074, "step": 90230 }, { "grad_norm": 0.09625670313835144, "learning_rate": 2.5823017899045565e-06, "loss": 0.0092, "step": 90240 }, { "grad_norm": 0.11869168281555176, "learning_rate": 2.577059344058047e-06, "loss": 0.0085, "step": 90250 }, { "grad_norm": 0.21115601062774658, "learning_rate": 2.5718220843130415e-06, "loss": 0.0095, "step": 90260 }, { "grad_norm": 0.2427811175584793, "learning_rate": 2.5665900112422535e-06, "loss": 0.0144, "step": 90270 }, { "grad_norm": 0.17309938371181488, "learning_rate": 2.561363125417876e-06, "loss": 0.0171, "step": 90280 }, { "grad_norm": 0.046324729919433594, "learning_rate": 2.556141427411496e-06, "loss": 0.0104, "step": 90290 }, { "grad_norm": 0.06780251860618591, "learning_rate": 2.550924917794156e-06, "loss": 0.0102, "step": 90300 }, { "grad_norm": 0.04552024230360985, "learning_rate": 2.545713597136329e-06, "loss": 0.0108, "step": 90310 }, { "grad_norm": 0.18155312538146973, "learning_rate": 2.5405074660079298e-06, "loss": 0.0085, "step": 90320 }, { "grad_norm": 0.06987904012203217, "learning_rate": 2.5353065249782647e-06, "loss": 0.007, "step": 90330 }, { "grad_norm": 0.20071662962436676, "learning_rate": 2.530110774616129e-06, "loss": 0.0137, "step": 90340 }, { "grad_norm": 0.07420634478330612, "learning_rate": 2.524920215489701e-06, "loss": 0.011, "step": 90350 }, { "grad_norm": 0.06675473600625992, "learning_rate": 2.519734848166627e-06, "loss": 0.0128, "step": 90360 }, { "grad_norm": 0.06658821552991867, "learning_rate": 2.514554673213948e-06, "loss": 0.0109, "step": 90370 }, { "grad_norm": 0.23010402917861938, "learning_rate": 2.5093796911981936e-06, "loss": 0.014, "step": 90380 }, { "grad_norm": 1.1924448013305664, "learning_rate": 2.5042099026852672e-06, "loss": 0.0116, "step": 90390 }, { "grad_norm": 0.07859372347593307, "learning_rate": 2.499045308240544e-06, "loss": 0.0085, "step": 90400 }, { "grad_norm": 0.21259130537509918, "learning_rate": 2.4938859084288003e-06, "loss": 0.0174, "step": 90410 }, { "grad_norm": 0.11736589670181274, "learning_rate": 2.4887317038142734e-06, "loss": 0.0084, "step": 90420 }, { "grad_norm": 0.12119589000940323, "learning_rate": 2.4835826949606124e-06, "loss": 0.01, "step": 90430 }, { "grad_norm": 0.2408890575170517, "learning_rate": 2.4784388824309167e-06, "loss": 0.0103, "step": 90440 }, { "grad_norm": 0.2065511792898178, "learning_rate": 2.473300266787687e-06, "loss": 0.0109, "step": 90450 }, { "grad_norm": 0.1793385148048401, "learning_rate": 2.46816684859289e-06, "loss": 0.0116, "step": 90460 }, { "grad_norm": 0.31227442622184753, "learning_rate": 2.4630386284079043e-06, "loss": 0.0105, "step": 90470 }, { "grad_norm": 0.11560288071632385, "learning_rate": 2.4579156067935427e-06, "loss": 0.0199, "step": 90480 }, { "grad_norm": 0.08309701085090637, "learning_rate": 2.452797784310046e-06, "loss": 0.0064, "step": 90490 }, { "grad_norm": 0.12213664501905441, "learning_rate": 2.447685161517105e-06, "loss": 0.0089, "step": 90500 }, { "grad_norm": 0.11744212359189987, "learning_rate": 2.4425777389738224e-06, "loss": 0.0107, "step": 90510 }, { "grad_norm": 0.11883704364299774, "learning_rate": 2.43747551723873e-06, "loss": 0.0101, "step": 90520 }, { "grad_norm": 0.11204897612333298, "learning_rate": 2.4323784968698136e-06, "loss": 0.012, "step": 90530 }, { "grad_norm": 0.1526743769645691, "learning_rate": 2.4272866784244565e-06, "loss": 0.0186, "step": 90540 }, { "grad_norm": 0.05259279906749725, "learning_rate": 2.422200062459518e-06, "loss": 0.0098, "step": 90550 }, { "grad_norm": 0.2462008148431778, "learning_rate": 2.41711864953123e-06, "loss": 0.0102, "step": 90560 }, { "grad_norm": 0.6399529576301575, "learning_rate": 2.412042440195322e-06, "loss": 0.0134, "step": 90570 }, { "grad_norm": 0.06693608313798904, "learning_rate": 2.4069714350068993e-06, "loss": 0.01, "step": 90580 }, { "grad_norm": 0.03170730918645859, "learning_rate": 2.4019056345205294e-06, "loss": 0.0059, "step": 90590 }, { "grad_norm": 0.37436532974243164, "learning_rate": 2.396845039290191e-06, "loss": 0.0124, "step": 90600 }, { "grad_norm": 0.18135273456573486, "learning_rate": 2.3917896498693194e-06, "loss": 0.0118, "step": 90610 }, { "grad_norm": 0.5067630410194397, "learning_rate": 2.3867394668107335e-06, "loss": 0.0139, "step": 90620 }, { "grad_norm": 0.1341976523399353, "learning_rate": 2.3816944906667527e-06, "loss": 0.01, "step": 90630 }, { "grad_norm": 0.15291672945022583, "learning_rate": 2.3766547219890634e-06, "loss": 0.0141, "step": 90640 }, { "grad_norm": 0.11185484379529953, "learning_rate": 2.371620161328819e-06, "loss": 0.0144, "step": 90650 }, { "grad_norm": 0.15970298647880554, "learning_rate": 2.366590809236574e-06, "loss": 0.0158, "step": 90660 }, { "grad_norm": 0.05911785364151001, "learning_rate": 2.3615666662623494e-06, "loss": 0.0077, "step": 90670 }, { "grad_norm": 0.08423870801925659, "learning_rate": 2.3565477329555664e-06, "loss": 0.0075, "step": 90680 }, { "grad_norm": 0.07135245203971863, "learning_rate": 2.3515340098650975e-06, "loss": 0.0104, "step": 90690 }, { "grad_norm": 0.5845035910606384, "learning_rate": 2.3465254975392258e-06, "loss": 0.014, "step": 90700 }, { "grad_norm": 0.26802605390548706, "learning_rate": 2.3415221965256805e-06, "loss": 0.0195, "step": 90710 }, { "grad_norm": 0.15641823410987854, "learning_rate": 2.3365241073716237e-06, "loss": 0.0141, "step": 90720 }, { "grad_norm": 0.397633820772171, "learning_rate": 2.3315312306236183e-06, "loss": 0.0093, "step": 90730 }, { "grad_norm": 0.15135744214057922, "learning_rate": 2.326543566827699e-06, "loss": 0.0136, "step": 90740 }, { "grad_norm": 0.1618906855583191, "learning_rate": 2.3215611165292862e-06, "loss": 0.0097, "step": 90750 }, { "grad_norm": 0.10861116647720337, "learning_rate": 2.3165838802732765e-06, "loss": 0.0089, "step": 90760 }, { "grad_norm": 0.14173175394535065, "learning_rate": 2.311611858603957e-06, "loss": 0.0077, "step": 90770 }, { "grad_norm": 0.13202838599681854, "learning_rate": 2.3066450520650763e-06, "loss": 0.0125, "step": 90780 }, { "grad_norm": 0.13083569705486298, "learning_rate": 2.3016834611997773e-06, "loss": 0.0231, "step": 90790 }, { "grad_norm": 0.5712804794311523, "learning_rate": 2.296727086550665e-06, "loss": 0.0154, "step": 90800 }, { "grad_norm": 0.12543989717960358, "learning_rate": 2.2917759286597507e-06, "loss": 0.0106, "step": 90810 }, { "grad_norm": 0.05788067728281021, "learning_rate": 2.2868299880685006e-06, "loss": 0.0095, "step": 90820 }, { "grad_norm": 0.18017050623893738, "learning_rate": 2.2818892653177827e-06, "loss": 0.0082, "step": 90830 }, { "grad_norm": 0.08869737386703491, "learning_rate": 2.2769537609479196e-06, "loss": 0.011, "step": 90840 }, { "grad_norm": 0.2983666658401489, "learning_rate": 2.2720234754986357e-06, "loss": 0.0119, "step": 90850 }, { "grad_norm": 0.13349926471710205, "learning_rate": 2.2670984095091108e-06, "loss": 0.0156, "step": 90860 }, { "grad_norm": 0.2422538548707962, "learning_rate": 2.262178563517936e-06, "loss": 0.019, "step": 90870 }, { "grad_norm": 0.26617857813835144, "learning_rate": 2.257263938063137e-06, "loss": 0.013, "step": 90880 }, { "grad_norm": 0.3383699357509613, "learning_rate": 2.2523545336821715e-06, "loss": 0.0158, "step": 90890 }, { "grad_norm": 0.14598320424556732, "learning_rate": 2.247450350911939e-06, "loss": 0.0159, "step": 90900 }, { "grad_norm": 0.23709701001644135, "learning_rate": 2.2425513902887328e-06, "loss": 0.0139, "step": 90910 }, { "grad_norm": 0.24818502366542816, "learning_rate": 2.2376576523483015e-06, "loss": 0.0136, "step": 90920 }, { "grad_norm": 0.11739073693752289, "learning_rate": 2.232769137625829e-06, "loss": 0.0121, "step": 90930 }, { "grad_norm": 0.27323877811431885, "learning_rate": 2.227885846655903e-06, "loss": 0.0127, "step": 90940 }, { "grad_norm": 0.34736841917037964, "learning_rate": 2.2230077799725523e-06, "loss": 0.011, "step": 90950 }, { "grad_norm": 0.3229336738586426, "learning_rate": 2.2181349381092396e-06, "loss": 0.0113, "step": 90960 }, { "grad_norm": 0.1683371216058731, "learning_rate": 2.2132673215988552e-06, "loss": 0.0129, "step": 90970 }, { "grad_norm": 0.15336178243160248, "learning_rate": 2.2084049309737065e-06, "loss": 0.0095, "step": 90980 }, { "grad_norm": 0.15043441951274872, "learning_rate": 2.2035477667655513e-06, "loss": 0.0094, "step": 90990 }, { "grad_norm": 0.14004594087600708, "learning_rate": 2.1986958295055315e-06, "loss": 0.0132, "step": 91000 }, { "grad_norm": 0.3542307913303375, "learning_rate": 2.193849119724284e-06, "loss": 0.0126, "step": 91010 }, { "grad_norm": 0.17855633795261383, "learning_rate": 2.1890076379518175e-06, "loss": 0.0096, "step": 91020 }, { "grad_norm": 0.021246083080768585, "learning_rate": 2.1841713847175925e-06, "loss": 0.0097, "step": 91030 }, { "grad_norm": 0.11591814458370209, "learning_rate": 2.1793403605504968e-06, "loss": 0.0145, "step": 91040 }, { "grad_norm": 0.16664984822273254, "learning_rate": 2.1745145659788413e-06, "loss": 0.0164, "step": 91050 }, { "grad_norm": 0.10705897957086563, "learning_rate": 2.1696940015303645e-06, "loss": 0.0074, "step": 91060 }, { "grad_norm": 0.15293680131435394, "learning_rate": 2.1648786677322397e-06, "loss": 0.0196, "step": 91070 }, { "grad_norm": 0.18408051133155823, "learning_rate": 2.1600685651110673e-06, "loss": 0.0074, "step": 91080 }, { "grad_norm": 0.24337142705917358, "learning_rate": 2.1552636941928717e-06, "loss": 0.0097, "step": 91090 }, { "grad_norm": 0.3343127369880676, "learning_rate": 2.1504640555031043e-06, "loss": 0.0098, "step": 91100 }, { "grad_norm": 0.18868659436702728, "learning_rate": 2.1456696495666506e-06, "loss": 0.0085, "step": 91110 }, { "grad_norm": 0.10308869183063507, "learning_rate": 2.1408804769078084e-06, "loss": 0.0129, "step": 91120 }, { "grad_norm": 0.23066963255405426, "learning_rate": 2.136096538050325e-06, "loss": 0.0219, "step": 91130 }, { "grad_norm": 0.261282354593277, "learning_rate": 2.13131783351736e-06, "loss": 0.0104, "step": 91140 }, { "grad_norm": 0.25947484374046326, "learning_rate": 2.126544363831512e-06, "loss": 0.0102, "step": 91150 }, { "grad_norm": 0.11634966731071472, "learning_rate": 2.121776129514791e-06, "loss": 0.0123, "step": 91160 }, { "grad_norm": 0.1924889087677002, "learning_rate": 2.1170131310886523e-06, "loss": 0.0086, "step": 91170 }, { "grad_norm": 0.08845969289541245, "learning_rate": 2.112255369073968e-06, "loss": 0.017, "step": 91180 }, { "grad_norm": 0.08414880186319351, "learning_rate": 2.1075028439910328e-06, "loss": 0.0097, "step": 91190 }, { "grad_norm": 0.1880367249250412, "learning_rate": 2.1027555563595813e-06, "loss": 0.0172, "step": 91200 }, { "grad_norm": 0.32133278250694275, "learning_rate": 2.0980135066987706e-06, "loss": 0.0089, "step": 91210 }, { "grad_norm": 0.15124735236167908, "learning_rate": 2.093276695527191e-06, "loss": 0.0116, "step": 91220 }, { "grad_norm": 0.09774543344974518, "learning_rate": 2.0885451233628396e-06, "loss": 0.0084, "step": 91230 }, { "grad_norm": 0.19146591424942017, "learning_rate": 2.0838187907231633e-06, "loss": 0.0097, "step": 91240 }, { "grad_norm": 0.3889620006084442, "learning_rate": 2.0790976981250153e-06, "loss": 0.0103, "step": 91250 }, { "grad_norm": 0.027805278077721596, "learning_rate": 2.0743818460847e-06, "loss": 0.0079, "step": 91260 }, { "grad_norm": 0.07593081891536713, "learning_rate": 2.069671235117926e-06, "loss": 0.0106, "step": 91270 }, { "grad_norm": 0.10698012262582779, "learning_rate": 2.0649658657398484e-06, "loss": 0.0126, "step": 91280 }, { "grad_norm": 0.06358049064874649, "learning_rate": 2.060265738465034e-06, "loss": 0.0071, "step": 91290 }, { "grad_norm": 0.08858942240476608, "learning_rate": 2.055570853807487e-06, "loss": 0.0082, "step": 91300 }, { "grad_norm": 0.19239018857479095, "learning_rate": 2.0508812122806196e-06, "loss": 0.0113, "step": 91310 }, { "grad_norm": 0.1359936147928238, "learning_rate": 2.046196814397294e-06, "loss": 0.0085, "step": 91320 }, { "grad_norm": 0.1374279260635376, "learning_rate": 2.0415176606697896e-06, "loss": 0.0125, "step": 91330 }, { "grad_norm": 0.11218397319316864, "learning_rate": 2.0368437516098126e-06, "loss": 0.009, "step": 91340 }, { "grad_norm": 0.12244763225317001, "learning_rate": 2.032175087728483e-06, "loss": 0.0119, "step": 91350 }, { "grad_norm": 0.09445950388908386, "learning_rate": 2.027511669536375e-06, "loss": 0.0136, "step": 91360 }, { "grad_norm": 0.10942326486110687, "learning_rate": 2.0228534975434643e-06, "loss": 0.0086, "step": 91370 }, { "grad_norm": 0.05767415463924408, "learning_rate": 2.0182005722591548e-06, "loss": 0.0118, "step": 91380 }, { "grad_norm": 0.08255191892385483, "learning_rate": 2.0135528941922944e-06, "loss": 0.0156, "step": 91390 }, { "grad_norm": 0.10097840428352356, "learning_rate": 2.008910463851149e-06, "loss": 0.0108, "step": 91400 }, { "grad_norm": 0.18227896094322205, "learning_rate": 2.0042732817433963e-06, "loss": 0.0096, "step": 91410 }, { "grad_norm": 0.28517457842826843, "learning_rate": 1.999641348376158e-06, "loss": 0.0176, "step": 91420 }, { "grad_norm": 0.16947196424007416, "learning_rate": 1.9950146642559787e-06, "loss": 0.012, "step": 91430 }, { "grad_norm": 0.06932423263788223, "learning_rate": 1.990393229888815e-06, "loss": 0.0082, "step": 91440 }, { "grad_norm": 0.2630104720592499, "learning_rate": 1.9857770457800684e-06, "loss": 0.0152, "step": 91450 }, { "grad_norm": 0.15494146943092346, "learning_rate": 1.9811661124345517e-06, "loss": 0.0145, "step": 91460 }, { "grad_norm": 0.21195709705352783, "learning_rate": 1.976560430356522e-06, "loss": 0.0083, "step": 91470 }, { "grad_norm": 0.44201090931892395, "learning_rate": 1.9719600000496383e-06, "loss": 0.0131, "step": 91480 }, { "grad_norm": 0.1405857354402542, "learning_rate": 1.967364822017004e-06, "loss": 0.0128, "step": 91490 }, { "grad_norm": 0.13135197758674622, "learning_rate": 1.962774896761127e-06, "loss": 0.0105, "step": 91500 }, { "grad_norm": 0.16839653253555298, "learning_rate": 1.958190224783968e-06, "loss": 0.0125, "step": 91510 }, { "grad_norm": 0.025474701076745987, "learning_rate": 1.9536108065868975e-06, "loss": 0.0062, "step": 91520 }, { "grad_norm": 0.030216051265597343, "learning_rate": 1.949036642670715e-06, "loss": 0.0074, "step": 91530 }, { "grad_norm": 0.12865957617759705, "learning_rate": 1.9444677335356377e-06, "loss": 0.015, "step": 91540 }, { "grad_norm": 0.15993061661720276, "learning_rate": 1.939904079681326e-06, "loss": 0.0168, "step": 91550 }, { "grad_norm": 0.23500823974609375, "learning_rate": 1.935345681606837e-06, "loss": 0.0103, "step": 91560 }, { "grad_norm": 0.17277146875858307, "learning_rate": 1.930792539810694e-06, "loss": 0.011, "step": 91570 }, { "grad_norm": 0.05873541161417961, "learning_rate": 1.9262446547907877e-06, "loss": 0.0132, "step": 91580 }, { "grad_norm": 0.12570521235466003, "learning_rate": 1.921702027044503e-06, "loss": 0.0094, "step": 91590 }, { "grad_norm": 0.18587292730808258, "learning_rate": 1.917164657068593e-06, "loss": 0.0101, "step": 91600 }, { "grad_norm": 0.06884802132844925, "learning_rate": 1.912632545359272e-06, "loss": 0.0117, "step": 91610 }, { "grad_norm": 0.24425847828388214, "learning_rate": 1.9081056924121488e-06, "loss": 0.0072, "step": 91620 }, { "grad_norm": 0.12997287511825562, "learning_rate": 1.9035840987222887e-06, "loss": 0.0113, "step": 91630 }, { "grad_norm": 0.15363258123397827, "learning_rate": 1.8990677647841515e-06, "loss": 0.0139, "step": 91640 }, { "grad_norm": 0.09269329160451889, "learning_rate": 1.8945566910916534e-06, "loss": 0.011, "step": 91650 }, { "grad_norm": 0.19109578430652618, "learning_rate": 1.8900508781381054e-06, "loss": 0.0117, "step": 91660 }, { "grad_norm": 0.054386161267757416, "learning_rate": 1.8855503264162633e-06, "loss": 0.0082, "step": 91670 }, { "grad_norm": 0.2567262649536133, "learning_rate": 1.8810550364182998e-06, "loss": 0.0117, "step": 91680 }, { "grad_norm": 0.04193153232336044, "learning_rate": 1.8765650086358112e-06, "loss": 0.0095, "step": 91690 }, { "grad_norm": 0.1932956427335739, "learning_rate": 1.8720802435598151e-06, "loss": 0.0145, "step": 91700 }, { "grad_norm": 0.2774308919906616, "learning_rate": 1.8676007416807694e-06, "loss": 0.0117, "step": 91710 }, { "grad_norm": 0.4658772349357605, "learning_rate": 1.8631265034885437e-06, "loss": 0.0135, "step": 91720 }, { "grad_norm": 0.05089588835835457, "learning_rate": 1.858657529472424e-06, "loss": 0.0062, "step": 91730 }, { "grad_norm": 0.33462658524513245, "learning_rate": 1.8541938201211473e-06, "loss": 0.0101, "step": 91740 }, { "grad_norm": 0.17131957411766052, "learning_rate": 1.8497353759228398e-06, "loss": 0.0115, "step": 91750 }, { "grad_norm": 0.08148709684610367, "learning_rate": 1.8452821973650835e-06, "loss": 0.0079, "step": 91760 }, { "grad_norm": 0.18151627480983734, "learning_rate": 1.840834284934856e-06, "loss": 0.0076, "step": 91770 }, { "grad_norm": 0.16884006559848785, "learning_rate": 1.8363916391185954e-06, "loss": 0.0086, "step": 91780 }, { "grad_norm": 0.10362719744443893, "learning_rate": 1.8319542604021245e-06, "loss": 0.0148, "step": 91790 }, { "grad_norm": 0.034589946269989014, "learning_rate": 1.8275221492707273e-06, "loss": 0.0053, "step": 91800 }, { "grad_norm": 0.3658825755119324, "learning_rate": 1.8230953062090717e-06, "loss": 0.0134, "step": 91810 }, { "grad_norm": 0.1264323592185974, "learning_rate": 1.8186737317012926e-06, "loss": 0.0058, "step": 91820 }, { "grad_norm": 0.11205897480249405, "learning_rate": 1.8142574262308977e-06, "loss": 0.0108, "step": 91830 }, { "grad_norm": 0.17046435177326202, "learning_rate": 1.8098463902808727e-06, "loss": 0.0144, "step": 91840 }, { "grad_norm": 0.3274444341659546, "learning_rate": 1.805440624333593e-06, "loss": 0.0068, "step": 91850 }, { "grad_norm": 0.09413035213947296, "learning_rate": 1.8010401288708734e-06, "loss": 0.0083, "step": 91860 }, { "grad_norm": 0.7183487415313721, "learning_rate": 1.7966449043739343e-06, "loss": 0.0099, "step": 91870 }, { "grad_norm": 0.22261197865009308, "learning_rate": 1.79225495132343e-06, "loss": 0.0146, "step": 91880 }, { "grad_norm": 0.276895672082901, "learning_rate": 1.7878702701994487e-06, "loss": 0.0192, "step": 91890 }, { "grad_norm": 0.04574303328990936, "learning_rate": 1.783490861481496e-06, "loss": 0.0103, "step": 91900 }, { "grad_norm": 0.20695434510707855, "learning_rate": 1.7791167256484832e-06, "loss": 0.0096, "step": 91910 }, { "grad_norm": 0.3828900456428528, "learning_rate": 1.774747863178766e-06, "loss": 0.0082, "step": 91920 }, { "grad_norm": 0.15151308476924896, "learning_rate": 1.7703842745501186e-06, "loss": 0.0126, "step": 91930 }, { "grad_norm": 0.1492663025856018, "learning_rate": 1.7660259602397366e-06, "loss": 0.006, "step": 91940 }, { "grad_norm": 0.2619643211364746, "learning_rate": 1.7616729207242388e-06, "loss": 0.0162, "step": 91950 }, { "grad_norm": 0.2233528196811676, "learning_rate": 1.7573251564796556e-06, "loss": 0.0095, "step": 91960 }, { "grad_norm": 0.4361661970615387, "learning_rate": 1.7529826679814676e-06, "loss": 0.0097, "step": 91970 }, { "grad_norm": 0.17584913969039917, "learning_rate": 1.7486454557045562e-06, "loss": 0.0102, "step": 91980 }, { "grad_norm": 0.15698471665382385, "learning_rate": 1.7443135201232364e-06, "loss": 0.0088, "step": 91990 }, { "grad_norm": 0.37428948283195496, "learning_rate": 1.7399868617112347e-06, "loss": 0.0175, "step": 92000 }, { "grad_norm": 0.22658658027648926, "learning_rate": 1.7356654809417171e-06, "loss": 0.0118, "step": 92010 }, { "grad_norm": 0.1467655599117279, "learning_rate": 1.7313493782872448e-06, "loss": 0.0158, "step": 92020 }, { "grad_norm": 0.1689101606607437, "learning_rate": 1.7270385542198453e-06, "loss": 0.0152, "step": 92030 }, { "grad_norm": 0.5481225848197937, "learning_rate": 1.7227330092109307e-06, "loss": 0.0105, "step": 92040 }, { "grad_norm": 0.1420501321554184, "learning_rate": 1.718432743731352e-06, "loss": 0.0067, "step": 92050 }, { "grad_norm": 0.05014714598655701, "learning_rate": 1.7141377582513718e-06, "loss": 0.0114, "step": 92060 }, { "grad_norm": 0.2348722219467163, "learning_rate": 1.709848053240698e-06, "loss": 0.011, "step": 92070 }, { "grad_norm": 0.16467827558517456, "learning_rate": 1.7055636291684328e-06, "loss": 0.0093, "step": 92080 }, { "grad_norm": 0.3183656334877014, "learning_rate": 1.7012844865031185e-06, "loss": 0.0132, "step": 92090 }, { "grad_norm": 0.08387922495603561, "learning_rate": 1.6970106257127138e-06, "loss": 0.0117, "step": 92100 }, { "grad_norm": 0.08574085682630539, "learning_rate": 1.6927420472646117e-06, "loss": 0.0079, "step": 92110 }, { "grad_norm": 0.09305078536272049, "learning_rate": 1.6884787516256106e-06, "loss": 0.01, "step": 92120 }, { "grad_norm": 0.07487396150827408, "learning_rate": 1.6842207392619325e-06, "loss": 0.0107, "step": 92130 }, { "grad_norm": 0.12211135774850845, "learning_rate": 1.679968010639238e-06, "loss": 0.0229, "step": 92140 }, { "grad_norm": 0.25548869371414185, "learning_rate": 1.6757205662225829e-06, "loss": 0.0118, "step": 92150 }, { "grad_norm": 0.0329812690615654, "learning_rate": 1.6714784064764844e-06, "loss": 0.0085, "step": 92160 }, { "grad_norm": 0.4354652762413025, "learning_rate": 1.6672415318648383e-06, "loss": 0.0108, "step": 92170 }, { "grad_norm": 0.17034193873405457, "learning_rate": 1.6630099428510016e-06, "loss": 0.0135, "step": 92180 }, { "grad_norm": 0.18550577759742737, "learning_rate": 1.6587836398977153e-06, "loss": 0.0161, "step": 92190 }, { "grad_norm": 0.08534053713083267, "learning_rate": 1.6545626234671762e-06, "loss": 0.0086, "step": 92200 }, { "grad_norm": 0.2810733914375305, "learning_rate": 1.650346894020971e-06, "loss": 0.0095, "step": 92210 }, { "grad_norm": 0.11298850923776627, "learning_rate": 1.6461364520201472e-06, "loss": 0.0068, "step": 92220 }, { "grad_norm": 0.187608003616333, "learning_rate": 1.6419312979251366e-06, "loss": 0.0168, "step": 92230 }, { "grad_norm": 0.0874461978673935, "learning_rate": 1.6377314321958214e-06, "loss": 0.0088, "step": 92240 }, { "grad_norm": 0.10100208967924118, "learning_rate": 1.6335368552914787e-06, "loss": 0.0095, "step": 92250 }, { "grad_norm": 0.3549751043319702, "learning_rate": 1.6293475676708358e-06, "loss": 0.0097, "step": 92260 }, { "grad_norm": 0.08417508006095886, "learning_rate": 1.6251635697920097e-06, "loss": 0.0165, "step": 92270 }, { "grad_norm": 0.30862244963645935, "learning_rate": 1.6209848621125624e-06, "loss": 0.0131, "step": 92280 }, { "grad_norm": 0.13925595581531525, "learning_rate": 1.6168114450894778e-06, "loss": 0.0124, "step": 92290 }, { "grad_norm": 0.14915220439434052, "learning_rate": 1.612643319179158e-06, "loss": 0.0084, "step": 92300 }, { "grad_norm": 0.4017433524131775, "learning_rate": 1.6084804848374046e-06, "loss": 0.0187, "step": 92310 }, { "grad_norm": 0.14071866869926453, "learning_rate": 1.6043229425194817e-06, "loss": 0.0154, "step": 92320 }, { "grad_norm": 0.137406125664711, "learning_rate": 1.6001706926800308e-06, "loss": 0.0123, "step": 92330 }, { "grad_norm": 0.13369640707969666, "learning_rate": 1.5960237357731444e-06, "loss": 0.0105, "step": 92340 }, { "grad_norm": 0.058744799345731735, "learning_rate": 1.5918820722523264e-06, "loss": 0.013, "step": 92350 }, { "grad_norm": 0.20002631843090057, "learning_rate": 1.587745702570509e-06, "loss": 0.0155, "step": 92360 }, { "grad_norm": 0.19462800025939941, "learning_rate": 1.5836146271800356e-06, "loss": 0.0132, "step": 92370 }, { "grad_norm": 0.19040435552597046, "learning_rate": 1.5794888465326728e-06, "loss": 0.0155, "step": 92380 }, { "grad_norm": 0.1795482635498047, "learning_rate": 1.5753683610796155e-06, "loss": 0.0121, "step": 92390 }, { "grad_norm": 0.09273218363523483, "learning_rate": 1.5712531712714528e-06, "loss": 0.0081, "step": 92400 }, { "grad_norm": 0.2899225950241089, "learning_rate": 1.5671432775582474e-06, "loss": 0.0122, "step": 92410 }, { "grad_norm": 0.11802420765161514, "learning_rate": 1.563038680389428e-06, "loss": 0.014, "step": 92420 }, { "grad_norm": 0.12986347079277039, "learning_rate": 1.5589393802138808e-06, "loss": 0.0112, "step": 92430 }, { "grad_norm": 0.16757604479789734, "learning_rate": 1.554845377479891e-06, "loss": 0.0135, "step": 92440 }, { "grad_norm": 0.26915982365608215, "learning_rate": 1.550756672635173e-06, "loss": 0.0154, "step": 92450 }, { "grad_norm": 0.17958302795886993, "learning_rate": 1.5466732661268635e-06, "loss": 0.0149, "step": 92460 }, { "grad_norm": 0.23709511756896973, "learning_rate": 1.5425951584015163e-06, "loss": 0.0141, "step": 92470 }, { "grad_norm": 0.10775548219680786, "learning_rate": 1.5385223499051026e-06, "loss": 0.026, "step": 92480 }, { "grad_norm": 0.12203626334667206, "learning_rate": 1.5344548410830328e-06, "loss": 0.0104, "step": 92490 }, { "grad_norm": 0.056716933846473694, "learning_rate": 1.5303926323801121e-06, "loss": 0.0079, "step": 92500 }, { "grad_norm": 0.15908312797546387, "learning_rate": 1.5263357242405795e-06, "loss": 0.0108, "step": 92510 }, { "grad_norm": 0.1766137033700943, "learning_rate": 1.5222841171080915e-06, "loss": 0.0113, "step": 92520 }, { "grad_norm": 0.23210395872592926, "learning_rate": 1.5182378114257268e-06, "loss": 0.0116, "step": 92530 }, { "grad_norm": 0.1485232263803482, "learning_rate": 1.5141968076359869e-06, "loss": 0.0067, "step": 92540 }, { "grad_norm": 0.05660216137766838, "learning_rate": 1.5101611061807908e-06, "loss": 0.0132, "step": 92550 }, { "grad_norm": 0.0712357759475708, "learning_rate": 1.506130707501463e-06, "loss": 0.0083, "step": 92560 }, { "grad_norm": 0.16966430842876434, "learning_rate": 1.5021056120387734e-06, "loss": 0.0124, "step": 92570 }, { "grad_norm": 0.16793741285800934, "learning_rate": 1.4980858202329084e-06, "loss": 0.0093, "step": 92580 }, { "grad_norm": 0.1380365639925003, "learning_rate": 1.4940713325234445e-06, "loss": 0.0238, "step": 92590 }, { "grad_norm": 0.09121402353048325, "learning_rate": 1.4900621493494137e-06, "loss": 0.0086, "step": 92600 }, { "grad_norm": 0.07197504490613937, "learning_rate": 1.4860582711492544e-06, "loss": 0.0123, "step": 92610 }, { "grad_norm": 0.11559759080410004, "learning_rate": 1.4820596983608325e-06, "loss": 0.0101, "step": 92620 }, { "grad_norm": 0.31425759196281433, "learning_rate": 1.4780664314214044e-06, "loss": 0.0142, "step": 92630 }, { "grad_norm": 0.07261467725038528, "learning_rate": 1.474078470767687e-06, "loss": 0.01, "step": 92640 }, { "grad_norm": 0.24194969236850739, "learning_rate": 1.470095816835787e-06, "loss": 0.0149, "step": 92650 }, { "grad_norm": 0.24585264921188354, "learning_rate": 1.4661184700612395e-06, "loss": 0.0087, "step": 92660 }, { "grad_norm": 0.06301688402891159, "learning_rate": 1.4621464308790133e-06, "loss": 0.0095, "step": 92670 }, { "grad_norm": 0.2725045084953308, "learning_rate": 1.4581796997234775e-06, "loss": 0.0104, "step": 92680 }, { "grad_norm": 0.11720301955938339, "learning_rate": 1.4542182770284297e-06, "loss": 0.008, "step": 92690 }, { "grad_norm": 0.3465202748775482, "learning_rate": 1.450262163227084e-06, "loss": 0.0105, "step": 92700 }, { "grad_norm": 0.08762291073799133, "learning_rate": 1.4463113587520671e-06, "loss": 0.0231, "step": 92710 }, { "grad_norm": 0.2774474620819092, "learning_rate": 1.4423658640354443e-06, "loss": 0.0129, "step": 92720 }, { "grad_norm": 0.1563338190317154, "learning_rate": 1.438425679508687e-06, "loss": 0.0185, "step": 92730 }, { "grad_norm": 0.06484927237033844, "learning_rate": 1.4344908056026895e-06, "loss": 0.007, "step": 92740 }, { "grad_norm": 0.06388358771800995, "learning_rate": 1.4305612427477521e-06, "loss": 0.009, "step": 92750 }, { "grad_norm": 0.2835580110549927, "learning_rate": 1.42663699137362e-06, "loss": 0.0088, "step": 92760 }, { "grad_norm": 0.10050450265407562, "learning_rate": 1.4227180519094386e-06, "loss": 0.0126, "step": 92770 }, { "grad_norm": 0.2033415585756302, "learning_rate": 1.4188044247837706e-06, "loss": 0.0149, "step": 92780 }, { "grad_norm": 0.09466582536697388, "learning_rate": 1.4148961104246128e-06, "loss": 0.01, "step": 92790 }, { "grad_norm": 0.07885514944791794, "learning_rate": 1.4109931092593731e-06, "loss": 0.0134, "step": 92800 }, { "grad_norm": 0.06947793811559677, "learning_rate": 1.4070954217148658e-06, "loss": 0.0101, "step": 92810 }, { "grad_norm": 0.045214105397462845, "learning_rate": 1.4032030482173497e-06, "loss": 0.0116, "step": 92820 }, { "grad_norm": 0.1254967302083969, "learning_rate": 1.3993159891924901e-06, "loss": 0.0097, "step": 92830 }, { "grad_norm": 0.21616023778915405, "learning_rate": 1.3954342450653524e-06, "loss": 0.0225, "step": 92840 }, { "grad_norm": 0.45360153913497925, "learning_rate": 1.3915578162604582e-06, "loss": 0.0161, "step": 92850 }, { "grad_norm": 0.057089392095804214, "learning_rate": 1.3876867032017126e-06, "loss": 0.0061, "step": 92860 }, { "grad_norm": 0.07237040251493454, "learning_rate": 1.383820906312472e-06, "loss": 0.0102, "step": 92870 }, { "grad_norm": 0.33202654123306274, "learning_rate": 1.3799604260154752e-06, "loss": 0.0165, "step": 92880 }, { "grad_norm": 0.27317580580711365, "learning_rate": 1.376105262732913e-06, "loss": 0.0155, "step": 92890 }, { "grad_norm": 0.15763521194458008, "learning_rate": 1.37225541688637e-06, "loss": 0.0145, "step": 92900 }, { "grad_norm": 0.4169010519981384, "learning_rate": 1.3684108888968594e-06, "loss": 0.0112, "step": 92910 }, { "grad_norm": 0.09532158821821213, "learning_rate": 1.364571679184823e-06, "loss": 0.0135, "step": 92920 }, { "grad_norm": 0.051739662885665894, "learning_rate": 1.3607377881701077e-06, "loss": 0.0164, "step": 92930 }, { "grad_norm": 0.05732153728604317, "learning_rate": 1.3569092162719787e-06, "loss": 0.0083, "step": 92940 }, { "grad_norm": 0.29920950531959534, "learning_rate": 1.353085963909123e-06, "loss": 0.0122, "step": 92950 }, { "grad_norm": 0.24329118430614471, "learning_rate": 1.3492680314996454e-06, "loss": 0.0093, "step": 92960 }, { "grad_norm": 0.07257161289453506, "learning_rate": 1.345455419461078e-06, "loss": 0.0151, "step": 92970 }, { "grad_norm": 0.08046606928110123, "learning_rate": 1.3416481282103431e-06, "loss": 0.0123, "step": 92980 }, { "grad_norm": 0.052837494760751724, "learning_rate": 1.337846158163819e-06, "loss": 0.0198, "step": 92990 }, { "grad_norm": 0.31351491808891296, "learning_rate": 1.334049509737273e-06, "loss": 0.0131, "step": 93000 }, { "grad_norm": 0.3636009991168976, "learning_rate": 1.3302581833459116e-06, "loss": 0.0112, "step": 93010 }, { "grad_norm": 0.028147229924798012, "learning_rate": 1.3264721794043366e-06, "loss": 0.0108, "step": 93020 }, { "grad_norm": 0.04578524827957153, "learning_rate": 1.3226914983265836e-06, "loss": 0.0118, "step": 93030 }, { "grad_norm": 0.07921145856380463, "learning_rate": 1.3189161405260997e-06, "loss": 0.0147, "step": 93040 }, { "grad_norm": 0.07760411500930786, "learning_rate": 1.3151461064157656e-06, "loss": 0.0125, "step": 93050 }, { "grad_norm": 0.046456266194581985, "learning_rate": 1.3113813964078459e-06, "loss": 0.0078, "step": 93060 }, { "grad_norm": 0.09531597793102264, "learning_rate": 1.3076220109140614e-06, "loss": 0.0114, "step": 93070 }, { "grad_norm": 0.031599462032318115, "learning_rate": 1.3038679503455275e-06, "loss": 0.0074, "step": 93080 }, { "grad_norm": 0.26638931035995483, "learning_rate": 1.3001192151127773e-06, "loss": 0.0102, "step": 93090 }, { "grad_norm": 0.04510956257581711, "learning_rate": 1.2963758056257657e-06, "loss": 0.0093, "step": 93100 }, { "grad_norm": 0.3662015497684479, "learning_rate": 1.2926377222938768e-06, "loss": 0.0134, "step": 93110 }, { "grad_norm": 0.26912614703178406, "learning_rate": 1.2889049655259e-06, "loss": 0.0197, "step": 93120 }, { "grad_norm": 0.3940277099609375, "learning_rate": 1.2851775357300312e-06, "loss": 0.0154, "step": 93130 }, { "grad_norm": 0.08188963681459427, "learning_rate": 1.281455433313916e-06, "loss": 0.0113, "step": 93140 }, { "grad_norm": 0.08871299773454666, "learning_rate": 1.2777386586845796e-06, "loss": 0.0056, "step": 93150 }, { "grad_norm": 0.05326131358742714, "learning_rate": 1.2740272122484964e-06, "loss": 0.0085, "step": 93160 }, { "grad_norm": 0.03005206771194935, "learning_rate": 1.2703210944115308e-06, "loss": 0.0094, "step": 93170 }, { "grad_norm": 0.25989651679992676, "learning_rate": 1.2666203055789916e-06, "loss": 0.0126, "step": 93180 }, { "grad_norm": 0.2824012041091919, "learning_rate": 1.2629248461555832e-06, "loss": 0.0143, "step": 93190 }, { "grad_norm": 0.27074065804481506, "learning_rate": 1.2592347165454431e-06, "loss": 0.0108, "step": 93200 }, { "grad_norm": 0.22072868049144745, "learning_rate": 1.2555499171521102e-06, "loss": 0.0085, "step": 93210 }, { "grad_norm": 0.19258509576320648, "learning_rate": 1.251870448378556e-06, "loss": 0.0115, "step": 93220 }, { "grad_norm": 0.1376439332962036, "learning_rate": 1.2481963106271478e-06, "loss": 0.0075, "step": 93230 }, { "grad_norm": 0.14001604914665222, "learning_rate": 1.2445275042997084e-06, "loss": 0.0107, "step": 93240 }, { "grad_norm": 0.06980367004871368, "learning_rate": 1.2408640297974283e-06, "loss": 0.0165, "step": 93250 }, { "grad_norm": 0.04839557409286499, "learning_rate": 1.2372058875209535e-06, "loss": 0.0111, "step": 93260 }, { "grad_norm": 0.038241948932409286, "learning_rate": 1.2335530778703252e-06, "loss": 0.0101, "step": 93270 }, { "grad_norm": 0.08696823567152023, "learning_rate": 1.2299056012450182e-06, "loss": 0.0156, "step": 93280 }, { "grad_norm": 0.38399577140808105, "learning_rate": 1.2262634580439026e-06, "loss": 0.0129, "step": 93290 }, { "grad_norm": 0.294391930103302, "learning_rate": 1.2226266486652926e-06, "loss": 0.0142, "step": 93300 }, { "grad_norm": 0.10538814961910248, "learning_rate": 1.2189951735068927e-06, "loss": 0.0083, "step": 93310 }, { "grad_norm": 0.23631204664707184, "learning_rate": 1.2153690329658407e-06, "loss": 0.0112, "step": 93320 }, { "grad_norm": 0.05074554309248924, "learning_rate": 1.2117482274386859e-06, "loss": 0.0096, "step": 93330 }, { "grad_norm": 0.11799739301204681, "learning_rate": 1.2081327573213953e-06, "loss": 0.0144, "step": 93340 }, { "grad_norm": 0.21280713379383087, "learning_rate": 1.2045226230093465e-06, "loss": 0.0106, "step": 93350 }, { "grad_norm": 0.12269818782806396, "learning_rate": 1.2009178248973296e-06, "loss": 0.0099, "step": 93360 }, { "grad_norm": 0.110792376101017, "learning_rate": 1.1973183633795849e-06, "loss": 0.0076, "step": 93370 }, { "grad_norm": 0.1020088791847229, "learning_rate": 1.193724238849725e-06, "loss": 0.0123, "step": 93380 }, { "grad_norm": 0.2313510775566101, "learning_rate": 1.190135451700808e-06, "loss": 0.0123, "step": 93390 }, { "grad_norm": 0.4355478286743164, "learning_rate": 1.186552002325292e-06, "loss": 0.0168, "step": 93400 }, { "grad_norm": 0.4670271873474121, "learning_rate": 1.1829738911150578e-06, "loss": 0.0083, "step": 93410 }, { "grad_norm": 0.1646905094385147, "learning_rate": 1.1794011184613985e-06, "loss": 0.0065, "step": 93420 }, { "grad_norm": 0.2438143640756607, "learning_rate": 1.17583368475504e-06, "loss": 0.0086, "step": 93430 }, { "grad_norm": 0.2173624038696289, "learning_rate": 1.1722715903860982e-06, "loss": 0.0077, "step": 93440 }, { "grad_norm": 0.12968626618385315, "learning_rate": 1.168714835744128e-06, "loss": 0.0044, "step": 93450 }, { "grad_norm": 0.24273283779621124, "learning_rate": 1.165163421218085e-06, "loss": 0.0155, "step": 93460 }, { "grad_norm": 0.10054740309715271, "learning_rate": 1.1616173471963477e-06, "loss": 0.0156, "step": 93470 }, { "grad_norm": 0.08673769235610962, "learning_rate": 1.1580766140667109e-06, "loss": 0.0101, "step": 93480 }, { "grad_norm": 0.15378910303115845, "learning_rate": 1.1545412222163764e-06, "loss": 0.0136, "step": 93490 }, { "grad_norm": 0.11371731758117676, "learning_rate": 1.1510111720319795e-06, "loss": 0.0179, "step": 93500 }, { "grad_norm": 0.04532487690448761, "learning_rate": 1.1474864638995663e-06, "loss": 0.0081, "step": 93510 }, { "grad_norm": 0.1394243687391281, "learning_rate": 1.1439670982045735e-06, "loss": 0.0135, "step": 93520 }, { "grad_norm": 0.07169999927282333, "learning_rate": 1.1404530753318877e-06, "loss": 0.0193, "step": 93530 }, { "grad_norm": 0.12918181717395782, "learning_rate": 1.136944395665801e-06, "loss": 0.0077, "step": 93540 }, { "grad_norm": 0.10411006957292557, "learning_rate": 1.1334410595900014e-06, "loss": 0.0056, "step": 93550 }, { "grad_norm": 0.5326501727104187, "learning_rate": 1.129943067487621e-06, "loss": 0.0152, "step": 93560 }, { "grad_norm": 0.11993404477834702, "learning_rate": 1.126450419741193e-06, "loss": 0.0159, "step": 93570 }, { "grad_norm": 0.2690180838108063, "learning_rate": 1.1229631167326671e-06, "loss": 0.0139, "step": 93580 }, { "grad_norm": 0.14286473393440247, "learning_rate": 1.1194811588434052e-06, "loss": 0.0104, "step": 93590 }, { "grad_norm": 0.0921163260936737, "learning_rate": 1.1160045464542023e-06, "loss": 0.0095, "step": 93600 }, { "grad_norm": 0.05702609941363335, "learning_rate": 1.1125332799452325e-06, "loss": 0.0074, "step": 93610 }, { "grad_norm": 0.34688955545425415, "learning_rate": 1.109067359696131e-06, "loss": 0.0171, "step": 93620 }, { "grad_norm": 0.26859164237976074, "learning_rate": 1.1056067860859166e-06, "loss": 0.0087, "step": 93630 }, { "grad_norm": 0.19470347464084625, "learning_rate": 1.102151559493031e-06, "loss": 0.0063, "step": 93640 }, { "grad_norm": 0.1509857028722763, "learning_rate": 1.0987016802953332e-06, "loss": 0.0122, "step": 93650 }, { "grad_norm": 0.07849431037902832, "learning_rate": 1.0952571488701048e-06, "loss": 0.0119, "step": 93660 }, { "grad_norm": 0.09222231060266495, "learning_rate": 1.0918179655940163e-06, "loss": 0.0133, "step": 93670 }, { "grad_norm": 0.12056279182434082, "learning_rate": 1.088384130843184e-06, "loss": 0.0105, "step": 93680 }, { "grad_norm": 0.4622654914855957, "learning_rate": 1.0849556449931296e-06, "loss": 0.0097, "step": 93690 }, { "grad_norm": 0.13001441955566406, "learning_rate": 1.0815325084187866e-06, "loss": 0.0101, "step": 93700 }, { "grad_norm": 0.09182162582874298, "learning_rate": 1.0781147214944943e-06, "loss": 0.0156, "step": 93710 }, { "grad_norm": 0.08221399784088135, "learning_rate": 1.0747022845940313e-06, "loss": 0.0104, "step": 93720 }, { "grad_norm": 0.05565229058265686, "learning_rate": 1.0712951980905605e-06, "loss": 0.0053, "step": 93730 }, { "grad_norm": 0.1943000853061676, "learning_rate": 1.0678934623566839e-06, "loss": 0.0172, "step": 93740 }, { "grad_norm": 0.09286490827798843, "learning_rate": 1.0644970777644091e-06, "loss": 0.0069, "step": 93750 }, { "grad_norm": 0.21946559846401215, "learning_rate": 1.0611060446851673e-06, "loss": 0.0105, "step": 93760 }, { "grad_norm": 0.11345305293798447, "learning_rate": 1.0577203634897836e-06, "loss": 0.0199, "step": 93770 }, { "grad_norm": 0.1733662188053131, "learning_rate": 1.0543400345485178e-06, "loss": 0.0111, "step": 93780 }, { "grad_norm": 0.030584707856178284, "learning_rate": 1.0509650582310404e-06, "loss": 0.0125, "step": 93790 }, { "grad_norm": 0.21566054224967957, "learning_rate": 1.0475954349064287e-06, "loss": 0.0061, "step": 93800 }, { "grad_norm": 0.08222386240959167, "learning_rate": 1.0442311649431825e-06, "loss": 0.0107, "step": 93810 }, { "grad_norm": 0.06644986569881439, "learning_rate": 1.0408722487092071e-06, "loss": 0.0128, "step": 93820 }, { "grad_norm": 0.059454597532749176, "learning_rate": 1.0375186865718424e-06, "loss": 0.0102, "step": 93830 }, { "grad_norm": 0.08716537803411484, "learning_rate": 1.0341704788978113e-06, "loss": 0.0123, "step": 93840 }, { "grad_norm": 0.22925682365894318, "learning_rate": 1.0308276260532879e-06, "loss": 0.0098, "step": 93850 }, { "grad_norm": 0.08350187540054321, "learning_rate": 1.0274901284038241e-06, "loss": 0.0046, "step": 93860 }, { "grad_norm": 0.07528393715620041, "learning_rate": 1.0241579863144114e-06, "loss": 0.0145, "step": 93870 }, { "grad_norm": 0.11974932998418808, "learning_rate": 1.0208312001494413e-06, "loss": 0.0083, "step": 93880 }, { "grad_norm": 0.26573455333709717, "learning_rate": 1.0175097702727453e-06, "loss": 0.0197, "step": 93890 }, { "grad_norm": 0.10916212201118469, "learning_rate": 1.0141936970475275e-06, "loss": 0.0058, "step": 93900 }, { "grad_norm": 0.06479103118181229, "learning_rate": 1.0108829808364417e-06, "loss": 0.0133, "step": 93910 }, { "grad_norm": 0.09646650403738022, "learning_rate": 1.0075776220015376e-06, "loss": 0.0125, "step": 93920 }, { "grad_norm": 0.22690866887569427, "learning_rate": 1.0042776209042816e-06, "loss": 0.015, "step": 93930 }, { "grad_norm": 0.07832136750221252, "learning_rate": 1.000982977905568e-06, "loss": 0.0109, "step": 93940 }, { "grad_norm": 0.4403727352619171, "learning_rate": 9.976936933656867e-07, "loss": 0.0073, "step": 93950 }, { "grad_norm": 0.18330737948417664, "learning_rate": 9.944097676443442e-07, "loss": 0.0137, "step": 93960 }, { "grad_norm": 0.542697548866272, "learning_rate": 9.911312011006757e-07, "loss": 0.0142, "step": 93970 }, { "grad_norm": 0.4485263526439667, "learning_rate": 9.878579940932109e-07, "loss": 0.0148, "step": 93980 }, { "grad_norm": 0.17797783017158508, "learning_rate": 9.84590146979908e-07, "loss": 0.0141, "step": 93990 }, { "grad_norm": 0.13895843923091888, "learning_rate": 9.813276601181365e-07, "loss": 0.0169, "step": 94000 }, { "grad_norm": 0.32851558923721313, "learning_rate": 9.780705338646724e-07, "loss": 0.0072, "step": 94010 }, { "grad_norm": 0.20818448066711426, "learning_rate": 9.748187685757083e-07, "loss": 0.0082, "step": 94020 }, { "grad_norm": 0.0582718588411808, "learning_rate": 9.715723646068543e-07, "loss": 0.0069, "step": 94030 }, { "grad_norm": 0.25897541642189026, "learning_rate": 9.68331322313143e-07, "loss": 0.0098, "step": 94040 }, { "grad_norm": 0.10136005282402039, "learning_rate": 9.650956420489909e-07, "loss": 0.0099, "step": 94050 }, { "grad_norm": 0.06864040344953537, "learning_rate": 9.618653241682595e-07, "loss": 0.0097, "step": 94060 }, { "grad_norm": 0.07599060982465744, "learning_rate": 9.586403690242053e-07, "loss": 0.01, "step": 94070 }, { "grad_norm": 0.040572360157966614, "learning_rate": 9.554207769695122e-07, "loss": 0.0128, "step": 94080 }, { "grad_norm": 0.1764921396970749, "learning_rate": 9.522065483562659e-07, "loss": 0.0112, "step": 94090 }, { "grad_norm": 0.1300155073404312, "learning_rate": 9.489976835359681e-07, "loss": 0.009, "step": 94100 }, { "grad_norm": 0.03315018117427826, "learning_rate": 9.457941828595329e-07, "loss": 0.0087, "step": 94110 }, { "grad_norm": 0.14240136742591858, "learning_rate": 9.425960466772965e-07, "loss": 0.0108, "step": 94120 }, { "grad_norm": 0.24454429745674133, "learning_rate": 9.394032753390014e-07, "loss": 0.0104, "step": 94130 }, { "grad_norm": 0.09581195563077927, "learning_rate": 9.362158691938017e-07, "loss": 0.0112, "step": 94140 }, { "grad_norm": 0.11256225407123566, "learning_rate": 9.330338285902685e-07, "loss": 0.008, "step": 94150 }, { "grad_norm": 0.2747635245323181, "learning_rate": 9.298571538763845e-07, "loss": 0.0113, "step": 94160 }, { "grad_norm": 0.21461153030395508, "learning_rate": 9.266858453995441e-07, "loss": 0.0126, "step": 94170 }, { "grad_norm": 0.08709625154733658, "learning_rate": 9.235199035065645e-07, "loss": 0.0103, "step": 94180 }, { "grad_norm": 0.0625138133764267, "learning_rate": 9.203593285436574e-07, "loss": 0.0089, "step": 94190 }, { "grad_norm": 0.09560166299343109, "learning_rate": 9.172041208564686e-07, "loss": 0.0096, "step": 94200 }, { "grad_norm": 0.10616730898618698, "learning_rate": 9.140542807900332e-07, "loss": 0.0098, "step": 94210 }, { "grad_norm": 0.17352260649204254, "learning_rate": 9.109098086888313e-07, "loss": 0.0096, "step": 94220 }, { "grad_norm": 0.09857923537492752, "learning_rate": 9.077707048967265e-07, "loss": 0.0111, "step": 94230 }, { "grad_norm": 0.08077696710824966, "learning_rate": 9.046369697570057e-07, "loss": 0.0124, "step": 94240 }, { "grad_norm": 0.23952177166938782, "learning_rate": 9.015086036123721e-07, "loss": 0.0149, "step": 94250 }, { "grad_norm": 0.16148175299167633, "learning_rate": 8.983856068049413e-07, "loss": 0.0079, "step": 94260 }, { "grad_norm": 0.44403520226478577, "learning_rate": 8.952679796762398e-07, "loss": 0.014, "step": 94270 }, { "grad_norm": 0.42277130484580994, "learning_rate": 8.921557225671951e-07, "loss": 0.0161, "step": 94280 }, { "grad_norm": 0.19500868022441864, "learning_rate": 8.890488358181792e-07, "loss": 0.0115, "step": 94290 }, { "grad_norm": 0.08051151782274246, "learning_rate": 8.859473197689372e-07, "loss": 0.0096, "step": 94300 }, { "grad_norm": 0.2703918218612671, "learning_rate": 8.828511747586588e-07, "loss": 0.007, "step": 94310 }, { "grad_norm": 0.2708846926689148, "learning_rate": 8.797604011259287e-07, "loss": 0.0087, "step": 94320 }, { "grad_norm": 0.12918445467948914, "learning_rate": 8.766749992087487e-07, "loss": 0.0058, "step": 94330 }, { "grad_norm": 0.26603588461875916, "learning_rate": 8.735949693445378e-07, "loss": 0.0091, "step": 94340 }, { "grad_norm": 0.10718738287687302, "learning_rate": 8.705203118701211e-07, "loss": 0.016, "step": 94350 }, { "grad_norm": 0.19476032257080078, "learning_rate": 8.674510271217295e-07, "loss": 0.0128, "step": 94360 }, { "grad_norm": 0.29655349254608154, "learning_rate": 8.643871154350336e-07, "loss": 0.0142, "step": 94370 }, { "grad_norm": 0.101972296833992, "learning_rate": 8.613285771450818e-07, "loss": 0.0146, "step": 94380 }, { "grad_norm": 0.07949525117874146, "learning_rate": 8.582754125863623e-07, "loss": 0.0081, "step": 94390 }, { "grad_norm": 0.167154923081398, "learning_rate": 8.55227622092758e-07, "loss": 0.0146, "step": 94400 }, { "grad_norm": 0.08992601186037064, "learning_rate": 8.521852059975743e-07, "loss": 0.0093, "step": 94410 }, { "grad_norm": 0.09776028245687485, "learning_rate": 8.491481646335231e-07, "loss": 0.0078, "step": 94420 }, { "grad_norm": 0.13221940398216248, "learning_rate": 8.46116498332733e-07, "loss": 0.0077, "step": 94430 }, { "grad_norm": 0.06033258140087128, "learning_rate": 8.430902074267333e-07, "loss": 0.0125, "step": 94440 }, { "grad_norm": 0.03948996216058731, "learning_rate": 8.400692922464925e-07, "loss": 0.0113, "step": 94450 }, { "grad_norm": 0.24130874872207642, "learning_rate": 8.370537531223577e-07, "loss": 0.0152, "step": 94460 }, { "grad_norm": 0.06604818254709244, "learning_rate": 8.340435903841092e-07, "loss": 0.0089, "step": 94470 }, { "grad_norm": 0.041495855897665024, "learning_rate": 8.310388043609341e-07, "loss": 0.0194, "step": 94480 }, { "grad_norm": 0.16884399950504303, "learning_rate": 8.280393953814303e-07, "loss": 0.0149, "step": 94490 }, { "grad_norm": 0.30407455563545227, "learning_rate": 8.250453637736077e-07, "loss": 0.0126, "step": 94500 }, { "grad_norm": 0.06270349770784378, "learning_rate": 8.22056709864899e-07, "loss": 0.0079, "step": 94510 }, { "grad_norm": 0.12936456501483917, "learning_rate": 8.190734339821204e-07, "loss": 0.0061, "step": 94520 }, { "grad_norm": 0.1849788874387741, "learning_rate": 8.160955364515333e-07, "loss": 0.0102, "step": 94530 }, { "grad_norm": 0.18271701037883759, "learning_rate": 8.131230175987936e-07, "loss": 0.013, "step": 94540 }, { "grad_norm": 0.18600976467132568, "learning_rate": 8.101558777489693e-07, "loss": 0.02, "step": 94550 }, { "grad_norm": 0.09050723910331726, "learning_rate": 8.071941172265396e-07, "loss": 0.0149, "step": 94560 }, { "grad_norm": 0.15789860486984253, "learning_rate": 8.042377363553955e-07, "loss": 0.0118, "step": 94570 }, { "grad_norm": 0.08732308447360992, "learning_rate": 8.01286735458856e-07, "loss": 0.0118, "step": 94580 }, { "grad_norm": 0.043530527502298355, "learning_rate": 7.983411148596298e-07, "loss": 0.012, "step": 94590 }, { "grad_norm": 0.32689058780670166, "learning_rate": 7.954008748798425e-07, "loss": 0.0176, "step": 94600 }, { "grad_norm": 0.08231793344020844, "learning_rate": 7.924660158410369e-07, "loss": 0.0117, "step": 94610 }, { "grad_norm": 0.3591819405555725, "learning_rate": 7.895365380641728e-07, "loss": 0.0155, "step": 94620 }, { "grad_norm": 0.2053525745868683, "learning_rate": 7.866124418695941e-07, "loss": 0.0105, "step": 94630 }, { "grad_norm": 0.28395217657089233, "learning_rate": 7.83693727577095e-07, "loss": 0.0123, "step": 94640 }, { "grad_norm": 0.16763097047805786, "learning_rate": 7.807803955058535e-07, "loss": 0.0103, "step": 94650 }, { "grad_norm": 0.04558452218770981, "learning_rate": 7.778724459744702e-07, "loss": 0.0106, "step": 94660 }, { "grad_norm": 0.14395740628242493, "learning_rate": 7.749698793009464e-07, "loss": 0.0111, "step": 94670 }, { "grad_norm": 0.5618685483932495, "learning_rate": 7.720726958027169e-07, "loss": 0.0087, "step": 94680 }, { "grad_norm": 0.03802783414721489, "learning_rate": 7.691808957965951e-07, "loss": 0.0055, "step": 94690 }, { "grad_norm": 0.17195497453212738, "learning_rate": 7.662944795988336e-07, "loss": 0.01, "step": 94700 }, { "grad_norm": 0.2248067855834961, "learning_rate": 7.634134475250854e-07, "loss": 0.0125, "step": 94710 }, { "grad_norm": 0.4253385365009308, "learning_rate": 7.605377998904262e-07, "loss": 0.0189, "step": 94720 }, { "grad_norm": 0.11468835175037384, "learning_rate": 7.5766753700931e-07, "loss": 0.0134, "step": 94730 }, { "grad_norm": 0.042992450296878815, "learning_rate": 7.54802659195647e-07, "loss": 0.0074, "step": 94740 }, { "grad_norm": 0.10073161125183105, "learning_rate": 7.519431667627253e-07, "loss": 0.0075, "step": 94750 }, { "grad_norm": 0.09027643501758575, "learning_rate": 7.490890600232503e-07, "loss": 0.0083, "step": 94760 }, { "grad_norm": 0.25559332966804504, "learning_rate": 7.462403392893502e-07, "loss": 0.0168, "step": 94770 }, { "grad_norm": 0.23042801022529602, "learning_rate": 7.433970048725536e-07, "loss": 0.0097, "step": 94780 }, { "grad_norm": 0.10624174773693085, "learning_rate": 7.405590570838062e-07, "loss": 0.0087, "step": 94790 }, { "grad_norm": 0.1857335865497589, "learning_rate": 7.377264962334596e-07, "loss": 0.0099, "step": 94800 }, { "grad_norm": 0.32232314348220825, "learning_rate": 7.348993226312773e-07, "loss": 0.0099, "step": 94810 }, { "grad_norm": 0.15782056748867035, "learning_rate": 7.32077536586434e-07, "loss": 0.0104, "step": 94820 }, { "grad_norm": 0.022935206070542336, "learning_rate": 7.292611384075221e-07, "loss": 0.0075, "step": 94830 }, { "grad_norm": 0.23910628259181976, "learning_rate": 7.264501284025338e-07, "loss": 0.0113, "step": 94840 }, { "grad_norm": 0.06554350256919861, "learning_rate": 7.236445068788789e-07, "loss": 0.0104, "step": 94850 }, { "grad_norm": 0.2845423221588135, "learning_rate": 7.20844274143373e-07, "loss": 0.0131, "step": 94860 }, { "grad_norm": 0.12409278750419617, "learning_rate": 7.180494305022489e-07, "loss": 0.0105, "step": 94870 }, { "grad_norm": 0.24755354225635529, "learning_rate": 7.152599762611456e-07, "loss": 0.0146, "step": 94880 }, { "grad_norm": 0.116141177713871, "learning_rate": 7.124759117251079e-07, "loss": 0.031, "step": 94890 }, { "grad_norm": 0.1502079963684082, "learning_rate": 7.096972371986033e-07, "loss": 0.0085, "step": 94900 }, { "grad_norm": 0.20406721532344818, "learning_rate": 7.069239529855054e-07, "loss": 0.0105, "step": 94910 }, { "grad_norm": 0.04688185453414917, "learning_rate": 7.041560593890939e-07, "loss": 0.0072, "step": 94920 }, { "grad_norm": 0.15571680665016174, "learning_rate": 7.013935567120599e-07, "loss": 0.0092, "step": 94930 }, { "grad_norm": 0.13627202808856964, "learning_rate": 6.986364452565009e-07, "loss": 0.0054, "step": 94940 }, { "grad_norm": 0.07937467843294144, "learning_rate": 6.95884725323942e-07, "loss": 0.0101, "step": 94950 }, { "grad_norm": 0.15160870552062988, "learning_rate": 6.931383972152982e-07, "loss": 0.0142, "step": 94960 }, { "grad_norm": 0.2656859755516052, "learning_rate": 6.903974612309127e-07, "loss": 0.0088, "step": 94970 }, { "grad_norm": 0.17997583746910095, "learning_rate": 6.876619176705179e-07, "loss": 0.0064, "step": 94980 }, { "grad_norm": 0.1179545521736145, "learning_rate": 6.8493176683328e-07, "loss": 0.0111, "step": 94990 }, { "grad_norm": 0.15511247515678406, "learning_rate": 6.822070090177657e-07, "loss": 0.0168, "step": 95000 }, { "grad_norm": 0.2173769772052765, "learning_rate": 6.794876445219367e-07, "loss": 0.0145, "step": 95010 }, { "grad_norm": 0.1620444655418396, "learning_rate": 6.767736736431885e-07, "loss": 0.0079, "step": 95020 }, { "grad_norm": 0.06350458413362503, "learning_rate": 6.740650966783113e-07, "loss": 0.0127, "step": 95030 }, { "grad_norm": 0.09248602390289307, "learning_rate": 6.713619139235239e-07, "loss": 0.0129, "step": 95040 }, { "grad_norm": 0.22237925231456757, "learning_rate": 6.686641256744286e-07, "loss": 0.0114, "step": 95050 }, { "grad_norm": 0.03225226700305939, "learning_rate": 6.659717322260562e-07, "loss": 0.0134, "step": 95060 }, { "grad_norm": 0.06431210786104202, "learning_rate": 6.632847338728377e-07, "loss": 0.0075, "step": 95070 }, { "grad_norm": 0.13872922956943512, "learning_rate": 6.60603130908627e-07, "loss": 0.0083, "step": 95080 }, { "grad_norm": 0.0676683783531189, "learning_rate": 6.579269236266783e-07, "loss": 0.012, "step": 95090 }, { "grad_norm": 0.35628634691238403, "learning_rate": 6.552561123196632e-07, "loss": 0.0079, "step": 95100 }, { "grad_norm": 0.4375455379486084, "learning_rate": 6.525906972796425e-07, "loss": 0.0163, "step": 95110 }, { "grad_norm": 0.07219842076301575, "learning_rate": 6.499306787981108e-07, "loss": 0.0057, "step": 95120 }, { "grad_norm": 0.40804651379585266, "learning_rate": 6.472760571659631e-07, "loss": 0.0094, "step": 95130 }, { "grad_norm": 0.3622884750366211, "learning_rate": 6.446268326735061e-07, "loss": 0.0145, "step": 95140 }, { "grad_norm": 0.1326223611831665, "learning_rate": 6.41983005610458e-07, "loss": 0.0097, "step": 95150 }, { "grad_norm": 0.07496193796396255, "learning_rate": 6.393445762659378e-07, "loss": 0.015, "step": 95160 }, { "grad_norm": 0.16860440373420715, "learning_rate": 6.36711544928481e-07, "loss": 0.0118, "step": 95170 }, { "grad_norm": 0.12446880340576172, "learning_rate": 6.340839118860298e-07, "loss": 0.0212, "step": 95180 }, { "grad_norm": 0.11802061647176743, "learning_rate": 6.314616774259485e-07, "loss": 0.01, "step": 95190 }, { "grad_norm": 0.05464082956314087, "learning_rate": 6.288448418349858e-07, "loss": 0.0036, "step": 95200 }, { "grad_norm": 0.04659207537770271, "learning_rate": 6.262334053993235e-07, "loss": 0.0158, "step": 95210 }, { "grad_norm": 0.28373539447784424, "learning_rate": 6.236273684045501e-07, "loss": 0.0103, "step": 95220 }, { "grad_norm": 0.12177536636590958, "learning_rate": 6.210267311356488e-07, "loss": 0.0103, "step": 95230 }, { "grad_norm": 0.18124118447303772, "learning_rate": 6.184314938770197e-07, "loss": 0.0093, "step": 95240 }, { "grad_norm": 0.35211044549942017, "learning_rate": 6.158416569124858e-07, "loss": 0.0097, "step": 95250 }, { "grad_norm": 0.13489215075969696, "learning_rate": 6.132572205252541e-07, "loss": 0.0078, "step": 95260 }, { "grad_norm": 0.18012851476669312, "learning_rate": 6.106781849979648e-07, "loss": 0.0124, "step": 95270 }, { "grad_norm": 0.08812634646892548, "learning_rate": 6.081045506126482e-07, "loss": 0.0144, "step": 95280 }, { "grad_norm": 0.2891550660133362, "learning_rate": 6.055363176507678e-07, "loss": 0.0184, "step": 95290 }, { "grad_norm": 0.060835301876068115, "learning_rate": 6.029734863931602e-07, "loss": 0.0147, "step": 95300 }, { "grad_norm": 0.3111453950405121, "learning_rate": 6.004160571201179e-07, "loss": 0.0155, "step": 95310 }, { "grad_norm": 0.09263347834348679, "learning_rate": 5.978640301112947e-07, "loss": 0.0123, "step": 95320 }, { "grad_norm": 0.20450827479362488, "learning_rate": 5.9531740564579e-07, "loss": 0.0142, "step": 95330 }, { "grad_norm": 0.24623045325279236, "learning_rate": 5.927761840020918e-07, "loss": 0.0086, "step": 95340 }, { "grad_norm": 0.025508752092719078, "learning_rate": 5.902403654581113e-07, "loss": 0.0088, "step": 95350 }, { "grad_norm": 0.13109086453914642, "learning_rate": 5.877099502911542e-07, "loss": 0.0144, "step": 95360 }, { "grad_norm": 0.183683842420578, "learning_rate": 5.851849387779496e-07, "loss": 0.0187, "step": 95370 }, { "grad_norm": 0.11831288039684296, "learning_rate": 5.826653311946262e-07, "loss": 0.0119, "step": 95380 }, { "grad_norm": 0.3023180663585663, "learning_rate": 5.801511278167249e-07, "loss": 0.015, "step": 95390 }, { "grad_norm": 0.06481605768203735, "learning_rate": 5.776423289191868e-07, "loss": 0.0099, "step": 95400 }, { "grad_norm": 0.2514830529689789, "learning_rate": 5.751389347763869e-07, "loss": 0.0129, "step": 95410 }, { "grad_norm": 0.3090008497238159, "learning_rate": 5.726409456620841e-07, "loss": 0.0145, "step": 95420 }, { "grad_norm": 0.18651001155376434, "learning_rate": 5.701483618494541e-07, "loss": 0.0151, "step": 95430 }, { "grad_norm": 0.1290394365787506, "learning_rate": 5.67661183611079e-07, "loss": 0.0149, "step": 95440 }, { "grad_norm": 0.07256340980529785, "learning_rate": 5.651794112189579e-07, "loss": 0.0099, "step": 95450 }, { "grad_norm": 0.08564333617687225, "learning_rate": 5.627030449444903e-07, "loss": 0.0103, "step": 95460 }, { "grad_norm": 0.3072335422039032, "learning_rate": 5.60232085058493e-07, "loss": 0.0131, "step": 95470 }, { "grad_norm": 0.023296313360333443, "learning_rate": 5.577665318311776e-07, "loss": 0.0104, "step": 95480 }, { "grad_norm": 0.056984782218933105, "learning_rate": 5.553063855321838e-07, "loss": 0.0117, "step": 95490 }, { "grad_norm": 0.24227957427501678, "learning_rate": 5.528516464305467e-07, "loss": 0.0133, "step": 95500 }, { "grad_norm": 0.08052144944667816, "learning_rate": 5.504023147947068e-07, "loss": 0.0129, "step": 95510 }, { "grad_norm": 0.07848044484853745, "learning_rate": 5.479583908925279e-07, "loss": 0.0137, "step": 95520 }, { "grad_norm": 0.09339767694473267, "learning_rate": 5.455198749912626e-07, "loss": 0.01, "step": 95530 }, { "grad_norm": 0.07000797986984253, "learning_rate": 5.430867673575979e-07, "loss": 0.0161, "step": 95540 }, { "grad_norm": 0.14630305767059326, "learning_rate": 5.406590682575985e-07, "loss": 0.0074, "step": 95550 }, { "grad_norm": 0.07598692178726196, "learning_rate": 5.382367779567687e-07, "loss": 0.0148, "step": 95560 }, { "grad_norm": 0.1770130693912506, "learning_rate": 5.358198967200023e-07, "loss": 0.0086, "step": 95570 }, { "grad_norm": 0.15394985675811768, "learning_rate": 5.334084248115989e-07, "loss": 0.0079, "step": 95580 }, { "grad_norm": 0.0538930669426918, "learning_rate": 5.310023624952754e-07, "loss": 0.0116, "step": 95590 }, { "grad_norm": 0.21363453567028046, "learning_rate": 5.286017100341656e-07, "loss": 0.0087, "step": 95600 }, { "grad_norm": 0.02582487463951111, "learning_rate": 5.262064676907874e-07, "loss": 0.0128, "step": 95610 }, { "grad_norm": 0.2799610197544098, "learning_rate": 5.238166357270924e-07, "loss": 0.0095, "step": 95620 }, { "grad_norm": 0.12354162335395813, "learning_rate": 5.214322144044215e-07, "loss": 0.011, "step": 95630 }, { "grad_norm": 0.11014352738857269, "learning_rate": 5.190532039835327e-07, "loss": 0.0154, "step": 95640 }, { "grad_norm": 0.11867818236351013, "learning_rate": 5.166796047245903e-07, "loss": 0.0111, "step": 95650 }, { "grad_norm": 0.13406983017921448, "learning_rate": 5.143114168871699e-07, "loss": 0.0093, "step": 95660 }, { "grad_norm": 0.21104870736598969, "learning_rate": 5.119486407302532e-07, "loss": 0.0113, "step": 95670 }, { "grad_norm": 0.11218655109405518, "learning_rate": 5.095912765122335e-07, "loss": 0.02, "step": 95680 }, { "grad_norm": 0.13978064060211182, "learning_rate": 5.07239324490899e-07, "loss": 0.0081, "step": 95690 }, { "grad_norm": 0.1680893450975418, "learning_rate": 5.048927849234608e-07, "loss": 0.0168, "step": 95700 }, { "grad_norm": 0.19817228615283966, "learning_rate": 5.025516580665358e-07, "loss": 0.0097, "step": 95710 }, { "grad_norm": 0.0670149028301239, "learning_rate": 5.002159441761412e-07, "loss": 0.0104, "step": 95720 }, { "grad_norm": 0.05310053005814552, "learning_rate": 4.978856435077062e-07, "loss": 0.0096, "step": 95730 }, { "grad_norm": 0.21142755448818207, "learning_rate": 4.955607563160769e-07, "loss": 0.0194, "step": 95740 }, { "grad_norm": 0.047589775174856186, "learning_rate": 4.932412828554944e-07, "loss": 0.0104, "step": 95750 }, { "grad_norm": 0.15292157232761383, "learning_rate": 4.909272233796114e-07, "loss": 0.0126, "step": 95760 }, { "grad_norm": 0.06251507252454758, "learning_rate": 4.88618578141492e-07, "loss": 0.0141, "step": 95770 }, { "grad_norm": 0.07165253162384033, "learning_rate": 4.863153473936011e-07, "loss": 0.0109, "step": 95780 }, { "grad_norm": 0.27686768770217896, "learning_rate": 4.840175313878259e-07, "loss": 0.0156, "step": 95790 }, { "grad_norm": 0.10645381361246109, "learning_rate": 4.817251303754433e-07, "loss": 0.0102, "step": 95800 }, { "grad_norm": 0.20926904678344727, "learning_rate": 4.794381446071583e-07, "loss": 0.0087, "step": 95810 }, { "grad_norm": 0.06753748655319214, "learning_rate": 4.771565743330597e-07, "loss": 0.0083, "step": 95820 }, { "grad_norm": 0.2288195937871933, "learning_rate": 4.7488041980266463e-07, "loss": 0.0098, "step": 95830 }, { "grad_norm": 0.16504137217998505, "learning_rate": 4.726096812648795e-07, "loss": 0.0164, "step": 95840 }, { "grad_norm": 0.41123273968696594, "learning_rate": 4.703443589680445e-07, "loss": 0.0135, "step": 95850 }, { "grad_norm": 0.12640975415706635, "learning_rate": 4.6808445315987826e-07, "loss": 0.0146, "step": 95860 }, { "grad_norm": 0.241448774933815, "learning_rate": 4.658299640875274e-07, "loss": 0.0122, "step": 95870 }, { "grad_norm": 0.13784992694854736, "learning_rate": 4.635808919975393e-07, "loss": 0.0095, "step": 95880 }, { "grad_norm": 1.1834206581115723, "learning_rate": 4.6133723713586706e-07, "loss": 0.0188, "step": 95890 }, { "grad_norm": 0.3056074380874634, "learning_rate": 4.590989997478756e-07, "loss": 0.0154, "step": 95900 }, { "grad_norm": 0.22742928564548492, "learning_rate": 4.5686618007833027e-07, "loss": 0.0142, "step": 95910 }, { "grad_norm": 0.08170183002948761, "learning_rate": 4.54638778371419e-07, "loss": 0.0115, "step": 95920 }, { "grad_norm": 0.05390952527523041, "learning_rate": 4.524167948707192e-07, "loss": 0.0124, "step": 95930 }, { "grad_norm": 0.4332648813724518, "learning_rate": 4.5020022981921986e-07, "loss": 0.0157, "step": 95940 }, { "grad_norm": 0.5240210890769958, "learning_rate": 4.47989083459327e-07, "loss": 0.0146, "step": 95950 }, { "grad_norm": 0.11186032742261887, "learning_rate": 4.457833560328528e-07, "loss": 0.0177, "step": 95960 }, { "grad_norm": 0.03952188417315483, "learning_rate": 4.4358304778100434e-07, "loss": 0.0089, "step": 95970 }, { "grad_norm": 0.24528272449970245, "learning_rate": 4.413881589444058e-07, "loss": 0.0199, "step": 95980 }, { "grad_norm": 0.06026705354452133, "learning_rate": 4.391986897630873e-07, "loss": 0.0156, "step": 95990 }, { "grad_norm": 0.08477889001369476, "learning_rate": 4.370146404764908e-07, "loss": 0.0113, "step": 96000 }, { "grad_norm": 0.07518114149570465, "learning_rate": 4.3483601132345285e-07, "loss": 0.0128, "step": 96010 }, { "grad_norm": 0.12802483141422272, "learning_rate": 4.32662802542233e-07, "loss": 0.0149, "step": 96020 }, { "grad_norm": 0.1753433346748352, "learning_rate": 4.3049501437047444e-07, "loss": 0.0073, "step": 96030 }, { "grad_norm": 0.05958957597613335, "learning_rate": 4.283326470452653e-07, "loss": 0.0156, "step": 96040 }, { "grad_norm": 0.1675814986228943, "learning_rate": 4.2617570080306647e-07, "loss": 0.0152, "step": 96050 }, { "grad_norm": 0.14907491207122803, "learning_rate": 4.2402417587975585e-07, "loss": 0.0112, "step": 96060 }, { "grad_norm": 0.05149566009640694, "learning_rate": 4.218780725106286e-07, "loss": 0.0078, "step": 96070 }, { "grad_norm": 0.1697642207145691, "learning_rate": 4.197373909303748e-07, "loss": 0.014, "step": 96080 }, { "grad_norm": 0.275081604719162, "learning_rate": 4.17602131373096e-07, "loss": 0.0074, "step": 96090 }, { "grad_norm": 0.056109923869371414, "learning_rate": 4.154722940723055e-07, "loss": 0.0215, "step": 96100 }, { "grad_norm": 0.054312389343976974, "learning_rate": 4.1334787926091136e-07, "loss": 0.0067, "step": 96110 }, { "grad_norm": 0.17232483625411987, "learning_rate": 4.1122888717124444e-07, "loss": 0.0088, "step": 96120 }, { "grad_norm": 0.23043543100357056, "learning_rate": 4.091153180350249e-07, "loss": 0.0108, "step": 96130 }, { "grad_norm": 0.21747995913028717, "learning_rate": 4.0700717208340125e-07, "loss": 0.0109, "step": 96140 }, { "grad_norm": 0.15203270316123962, "learning_rate": 4.049044495469112e-07, "loss": 0.0083, "step": 96150 }, { "grad_norm": 0.041893329471349716, "learning_rate": 4.0280715065549866e-07, "loss": 0.008, "step": 96160 }, { "grad_norm": 0.0992785096168518, "learning_rate": 4.0071527563853575e-07, "loss": 0.0142, "step": 96170 }, { "grad_norm": 0.043958913534879684, "learning_rate": 3.9862882472477823e-07, "loss": 0.01, "step": 96180 }, { "grad_norm": 0.12668052315711975, "learning_rate": 3.9654779814239375e-07, "loss": 0.0082, "step": 96190 }, { "grad_norm": 0.04977337643504143, "learning_rate": 3.944721961189668e-07, "loss": 0.0066, "step": 96200 }, { "grad_norm": 0.05939456820487976, "learning_rate": 3.9240201888147697e-07, "loss": 0.0113, "step": 96210 }, { "grad_norm": 0.05126618221402168, "learning_rate": 3.9033726665632096e-07, "loss": 0.0186, "step": 96220 }, { "grad_norm": 0.3095126450061798, "learning_rate": 3.882779396692959e-07, "loss": 0.0132, "step": 96230 }, { "grad_norm": 0.12885235249996185, "learning_rate": 3.8622403814559947e-07, "loss": 0.0082, "step": 96240 }, { "grad_norm": 0.23340336978435516, "learning_rate": 3.841755623098575e-07, "loss": 0.0081, "step": 96250 }, { "grad_norm": 0.19126677513122559, "learning_rate": 3.821325123860742e-07, "loss": 0.0083, "step": 96260 }, { "grad_norm": 0.3229627013206482, "learning_rate": 3.800948885976818e-07, "loss": 0.0116, "step": 96270 }, { "grad_norm": 0.06531715393066406, "learning_rate": 3.780626911675078e-07, "loss": 0.0115, "step": 96280 }, { "grad_norm": 0.0737352967262268, "learning_rate": 3.76035920317791e-07, "loss": 0.0122, "step": 96290 }, { "grad_norm": 0.05921710282564163, "learning_rate": 3.740145762701819e-07, "loss": 0.0084, "step": 96300 }, { "grad_norm": 0.08542086184024811, "learning_rate": 3.7199865924572585e-07, "loss": 0.0085, "step": 96310 }, { "grad_norm": 0.15893736481666565, "learning_rate": 3.699881694648799e-07, "loss": 0.0066, "step": 96320 }, { "grad_norm": 0.31897687911987305, "learning_rate": 3.679831071475126e-07, "loss": 0.0124, "step": 96330 }, { "grad_norm": 0.09968569874763489, "learning_rate": 3.65983472512893e-07, "loss": 0.0207, "step": 96340 }, { "grad_norm": 0.07753830403089523, "learning_rate": 3.6398926577969616e-07, "loss": 0.0076, "step": 96350 }, { "grad_norm": 0.03609111160039902, "learning_rate": 3.620004871660032e-07, "loss": 0.0101, "step": 96360 }, { "grad_norm": 0.042845360934734344, "learning_rate": 3.600171368893124e-07, "loss": 0.0071, "step": 96370 }, { "grad_norm": 0.06957073509693146, "learning_rate": 3.580392151665113e-07, "loss": 0.0078, "step": 96380 }, { "grad_norm": 0.058295950293540955, "learning_rate": 3.5606672221391024e-07, "loss": 0.0091, "step": 96390 }, { "grad_norm": 0.255414217710495, "learning_rate": 3.540996582472089e-07, "loss": 0.0132, "step": 96400 }, { "grad_norm": 0.13031898438930511, "learning_rate": 3.521380234815297e-07, "loss": 0.0117, "step": 96410 }, { "grad_norm": 0.21634341776371002, "learning_rate": 3.5018181813138985e-07, "loss": 0.0111, "step": 96420 }, { "grad_norm": 0.3064461052417755, "learning_rate": 3.482310424107238e-07, "loss": 0.0106, "step": 96430 }, { "grad_norm": 0.10142865777015686, "learning_rate": 3.4628569653285537e-07, "loss": 0.0122, "step": 96440 }, { "grad_norm": 0.06816349178552628, "learning_rate": 3.44345780710531e-07, "loss": 0.0116, "step": 96450 }, { "grad_norm": 0.41840478777885437, "learning_rate": 3.424112951558978e-07, "loss": 0.0167, "step": 96460 }, { "grad_norm": 0.11046569049358368, "learning_rate": 3.40482240080503e-07, "loss": 0.0148, "step": 96470 }, { "grad_norm": 0.2057236135005951, "learning_rate": 3.3855861569531133e-07, "loss": 0.0133, "step": 96480 }, { "grad_norm": 0.34343940019607544, "learning_rate": 3.3664042221068226e-07, "loss": 0.0146, "step": 96490 }, { "grad_norm": 0.5068454742431641, "learning_rate": 3.347276598363924e-07, "loss": 0.0104, "step": 96500 }, { "grad_norm": 0.06612684577703476, "learning_rate": 3.328203287816134e-07, "loss": 0.0066, "step": 96510 }, { "grad_norm": 0.09376511722803116, "learning_rate": 3.3091842925493386e-07, "loss": 0.0163, "step": 96520 }, { "grad_norm": 0.1241193488240242, "learning_rate": 3.2902196146433196e-07, "loss": 0.0085, "step": 96530 }, { "grad_norm": 0.10442811250686646, "learning_rate": 3.2713092561720836e-07, "loss": 0.0119, "step": 96540 }, { "grad_norm": 0.2917722165584564, "learning_rate": 3.2524532192036996e-07, "loss": 0.0074, "step": 96550 }, { "grad_norm": 0.13488087058067322, "learning_rate": 3.2336515058001836e-07, "loss": 0.0105, "step": 96560 }, { "grad_norm": 0.1393156200647354, "learning_rate": 3.2149041180176143e-07, "loss": 0.0097, "step": 96570 }, { "grad_norm": 0.06165351718664169, "learning_rate": 3.1962110579062954e-07, "loss": 0.0076, "step": 96580 }, { "grad_norm": 0.09925521910190582, "learning_rate": 3.177572327510314e-07, "loss": 0.0055, "step": 96590 }, { "grad_norm": 0.25198206305503845, "learning_rate": 3.158987928868151e-07, "loss": 0.0154, "step": 96600 }, { "grad_norm": 0.08116254210472107, "learning_rate": 3.1404578640120144e-07, "loss": 0.0122, "step": 96610 }, { "grad_norm": 0.09970688074827194, "learning_rate": 3.1219821349683933e-07, "loss": 0.0107, "step": 96620 }, { "grad_norm": 0.10418365150690079, "learning_rate": 3.103560743757783e-07, "loss": 0.0124, "step": 96630 }, { "grad_norm": 0.08932279050350189, "learning_rate": 3.0851936923946834e-07, "loss": 0.0104, "step": 96640 }, { "grad_norm": 0.09717793762683868, "learning_rate": 3.06688098288771e-07, "loss": 0.0088, "step": 96650 }, { "grad_norm": 0.3373450040817261, "learning_rate": 3.0486226172394825e-07, "loss": 0.0114, "step": 96660 }, { "grad_norm": 0.05057382956147194, "learning_rate": 3.030418597446738e-07, "loss": 0.0101, "step": 96670 }, { "grad_norm": 0.0812063068151474, "learning_rate": 3.012268925500272e-07, "loss": 0.0123, "step": 96680 }, { "grad_norm": 0.19797289371490479, "learning_rate": 2.994173603384831e-07, "loss": 0.012, "step": 96690 }, { "grad_norm": 0.08255276829004288, "learning_rate": 2.9761326330793317e-07, "loss": 0.0091, "step": 96700 }, { "grad_norm": 0.2642654776573181, "learning_rate": 2.958146016556751e-07, "loss": 0.0128, "step": 96710 }, { "grad_norm": 0.03108764998614788, "learning_rate": 2.94021375578396e-07, "loss": 0.0089, "step": 96720 }, { "grad_norm": 0.04956062510609627, "learning_rate": 2.9223358527221135e-07, "loss": 0.0093, "step": 96730 }, { "grad_norm": 0.05594097077846527, "learning_rate": 2.904512309326257e-07, "loss": 0.0137, "step": 96740 }, { "grad_norm": 0.2533566951751709, "learning_rate": 2.8867431275455545e-07, "loss": 0.0104, "step": 96750 }, { "grad_norm": 0.10367424786090851, "learning_rate": 2.86902830932323e-07, "loss": 0.0119, "step": 96760 }, { "grad_norm": 0.19294849038124084, "learning_rate": 2.851367856596565e-07, "loss": 0.0121, "step": 96770 }, { "grad_norm": 0.10175339877605438, "learning_rate": 2.83376177129685e-07, "loss": 0.0142, "step": 96780 }, { "grad_norm": 0.2957955598831177, "learning_rate": 2.8162100553494884e-07, "loss": 0.0123, "step": 96790 }, { "grad_norm": 0.3180539906024933, "learning_rate": 2.798712710673834e-07, "loss": 0.0111, "step": 96800 }, { "grad_norm": 0.1171446293592453, "learning_rate": 2.7812697391834676e-07, "loss": 0.0088, "step": 96810 }, { "grad_norm": 0.06576794385910034, "learning_rate": 2.7638811427858627e-07, "loss": 0.0094, "step": 96820 }, { "grad_norm": 0.19100503623485565, "learning_rate": 2.746546923382665e-07, "loss": 0.011, "step": 96830 }, { "grad_norm": 0.051207926124334335, "learning_rate": 2.729267082869469e-07, "loss": 0.0142, "step": 96840 }, { "grad_norm": 0.0616631843149662, "learning_rate": 2.712041623135986e-07, "loss": 0.0139, "step": 96850 }, { "grad_norm": 0.12747183442115784, "learning_rate": 2.694870546065931e-07, "loss": 0.0115, "step": 96860 }, { "grad_norm": 0.4996401369571686, "learning_rate": 2.677753853537246e-07, "loss": 0.0109, "step": 96870 }, { "grad_norm": 0.07401274144649506, "learning_rate": 2.660691547421601e-07, "loss": 0.0145, "step": 96880 }, { "grad_norm": 0.025067556649446487, "learning_rate": 2.6436836295850033e-07, "loss": 0.0158, "step": 96890 }, { "grad_norm": 0.08242125064134598, "learning_rate": 2.6267301018874093e-07, "loss": 0.0153, "step": 96900 }, { "grad_norm": 0.07966786623001099, "learning_rate": 2.6098309661828355e-07, "loss": 0.0125, "step": 96910 }, { "grad_norm": 0.4825231730937958, "learning_rate": 2.592986224319305e-07, "loss": 0.0089, "step": 96920 }, { "grad_norm": 0.10385879874229431, "learning_rate": 2.576195878139009e-07, "loss": 0.0121, "step": 96930 }, { "grad_norm": 0.06522275507450104, "learning_rate": 2.5594599294780364e-07, "loss": 0.0086, "step": 96940 }, { "grad_norm": 0.31438979506492615, "learning_rate": 2.542778380166644e-07, "loss": 0.0134, "step": 96950 }, { "grad_norm": 0.18955841660499573, "learning_rate": 2.526151232029095e-07, "loss": 0.0119, "step": 96960 }, { "grad_norm": 0.3067888915538788, "learning_rate": 2.5095784868836567e-07, "loss": 0.0118, "step": 96970 }, { "grad_norm": 0.16088762879371643, "learning_rate": 2.493060146542825e-07, "loss": 0.0066, "step": 96980 }, { "grad_norm": 0.1425389051437378, "learning_rate": 2.476596212812876e-07, "loss": 0.0075, "step": 96990 }, { "grad_norm": 0.46088820695877075, "learning_rate": 2.4601866874943703e-07, "loss": 0.012, "step": 97000 }, { "grad_norm": 0.26781192421913147, "learning_rate": 2.4438315723818163e-07, "loss": 0.0097, "step": 97010 }, { "grad_norm": 0.21955260634422302, "learning_rate": 2.4275308692637833e-07, "loss": 0.009, "step": 97020 }, { "grad_norm": 0.1268150955438614, "learning_rate": 2.411284579922901e-07, "loss": 0.0136, "step": 97030 }, { "grad_norm": 0.39134928584098816, "learning_rate": 2.395092706135804e-07, "loss": 0.0173, "step": 97040 }, { "grad_norm": 0.17431794106960297, "learning_rate": 2.3789552496731872e-07, "loss": 0.0063, "step": 97050 }, { "grad_norm": 0.26291266083717346, "learning_rate": 2.3628722122999158e-07, "loss": 0.0081, "step": 97060 }, { "grad_norm": 0.22919587790966034, "learning_rate": 2.3468435957747503e-07, "loss": 0.0114, "step": 97070 }, { "grad_norm": 0.18062958121299744, "learning_rate": 2.330869401850566e-07, "loss": 0.0124, "step": 97080 }, { "grad_norm": 0.050829604268074036, "learning_rate": 2.3149496322742437e-07, "loss": 0.0088, "step": 97090 }, { "grad_norm": 0.09909895807504654, "learning_rate": 2.2990842887868346e-07, "loss": 0.0146, "step": 97100 }, { "grad_norm": 0.07455948740243912, "learning_rate": 2.2832733731232292e-07, "loss": 0.0099, "step": 97110 }, { "grad_norm": 0.08314575254917145, "learning_rate": 2.2675168870124886e-07, "loss": 0.0078, "step": 97120 }, { "grad_norm": 0.210254967212677, "learning_rate": 2.2518148321778455e-07, "loss": 0.0101, "step": 97130 }, { "grad_norm": 0.1678897738456726, "learning_rate": 2.2361672103363706e-07, "loss": 0.0135, "step": 97140 }, { "grad_norm": 0.08507867902517319, "learning_rate": 2.220574023199251e-07, "loss": 0.0111, "step": 97150 }, { "grad_norm": 0.11885147541761398, "learning_rate": 2.2050352724717894e-07, "loss": 0.008, "step": 97160 }, { "grad_norm": 0.19009488821029663, "learning_rate": 2.1895509598532372e-07, "loss": 0.0099, "step": 97170 }, { "grad_norm": 0.30296334624290466, "learning_rate": 2.1741210870369066e-07, "loss": 0.0102, "step": 97180 }, { "grad_norm": 0.07716668397188187, "learning_rate": 2.1587456557102814e-07, "loss": 0.0111, "step": 97190 }, { "grad_norm": 0.07409197837114334, "learning_rate": 2.143424667554683e-07, "loss": 0.0098, "step": 97200 }, { "grad_norm": 0.3212556838989258, "learning_rate": 2.128158124245716e-07, "loss": 0.0125, "step": 97210 }, { "grad_norm": 0.10185685008764267, "learning_rate": 2.1129460274527668e-07, "loss": 0.0095, "step": 97220 }, { "grad_norm": 0.38554948568344116, "learning_rate": 2.0977883788395046e-07, "loss": 0.0068, "step": 97230 }, { "grad_norm": 0.05086122080683708, "learning_rate": 2.0826851800634928e-07, "loss": 0.0099, "step": 97240 }, { "grad_norm": 0.1479480266571045, "learning_rate": 2.0676364327764653e-07, "loss": 0.0122, "step": 97250 }, { "grad_norm": 0.17042914032936096, "learning_rate": 2.0526421386240502e-07, "loss": 0.0116, "step": 97260 }, { "grad_norm": 0.11750880628824234, "learning_rate": 2.0377022992461025e-07, "loss": 0.0146, "step": 97270 }, { "grad_norm": 0.11063122004270554, "learning_rate": 2.0228169162763156e-07, "loss": 0.0113, "step": 97280 }, { "grad_norm": 0.11539747565984726, "learning_rate": 2.0079859913425536e-07, "loss": 0.0148, "step": 97290 }, { "grad_norm": 0.13784611225128174, "learning_rate": 1.9932095260667417e-07, "loss": 0.0131, "step": 97300 }, { "grad_norm": 0.11026003211736679, "learning_rate": 1.9784875220648092e-07, "loss": 0.0107, "step": 97310 }, { "grad_norm": 0.07803951948881149, "learning_rate": 1.9638199809466907e-07, "loss": 0.0084, "step": 97320 }, { "grad_norm": 0.1705700010061264, "learning_rate": 1.9492069043164364e-07, "loss": 0.0148, "step": 97330 }, { "grad_norm": 0.19135689735412598, "learning_rate": 1.9346482937721567e-07, "loss": 0.0123, "step": 97340 }, { "grad_norm": 0.22170519828796387, "learning_rate": 1.920144150905856e-07, "loss": 0.012, "step": 97350 }, { "grad_norm": 0.2344105988740921, "learning_rate": 1.9056944773037655e-07, "loss": 0.0167, "step": 97360 }, { "grad_norm": 0.12845966219902039, "learning_rate": 1.8912992745460102e-07, "loss": 0.0088, "step": 97370 }, { "grad_norm": 0.22927676141262054, "learning_rate": 1.876958544206886e-07, "loss": 0.017, "step": 97380 }, { "grad_norm": 0.11173975467681885, "learning_rate": 1.8626722878546942e-07, "loss": 0.0139, "step": 97390 }, { "grad_norm": 0.39573732018470764, "learning_rate": 1.8484405070516851e-07, "loss": 0.0076, "step": 97400 }, { "grad_norm": 0.2782362103462219, "learning_rate": 1.8342632033542805e-07, "loss": 0.0142, "step": 97410 }, { "grad_norm": 0.15123452246189117, "learning_rate": 1.8201403783129066e-07, "loss": 0.0119, "step": 97420 }, { "grad_norm": 0.14394479990005493, "learning_rate": 1.806072033471884e-07, "loss": 0.0101, "step": 97430 }, { "grad_norm": 0.21302595734596252, "learning_rate": 1.79205817036987e-07, "loss": 0.0197, "step": 97440 }, { "grad_norm": 0.07234370708465576, "learning_rate": 1.7780987905393065e-07, "loss": 0.0112, "step": 97450 }, { "grad_norm": 0.22502276301383972, "learning_rate": 1.764193895506805e-07, "loss": 0.0105, "step": 97460 }, { "grad_norm": 0.33382996916770935, "learning_rate": 1.7503434867929824e-07, "loss": 0.0119, "step": 97470 }, { "grad_norm": 0.1253911405801773, "learning_rate": 1.7365475659125165e-07, "loss": 0.0089, "step": 97480 }, { "grad_norm": 0.19342990219593048, "learning_rate": 1.7228061343740332e-07, "loss": 0.0139, "step": 97490 }, { "grad_norm": 0.42956164479255676, "learning_rate": 1.7091191936803308e-07, "loss": 0.0098, "step": 97500 }, { "grad_norm": 0.18197178840637207, "learning_rate": 1.6954867453281563e-07, "loss": 0.0071, "step": 97510 }, { "grad_norm": 0.19924356043338776, "learning_rate": 1.6819087908084285e-07, "loss": 0.0156, "step": 97520 }, { "grad_norm": 0.1743813157081604, "learning_rate": 1.6683853316058485e-07, "loss": 0.01, "step": 97530 }, { "grad_norm": 0.14488983154296875, "learning_rate": 1.6549163691995106e-07, "loss": 0.0132, "step": 97540 }, { "grad_norm": 0.2269967645406723, "learning_rate": 1.641501905062237e-07, "loss": 0.0254, "step": 97550 }, { "grad_norm": 0.12345749884843826, "learning_rate": 1.628141940661021e-07, "loss": 0.0164, "step": 97560 }, { "grad_norm": 0.09143583476543427, "learning_rate": 1.6148364774569158e-07, "loss": 0.0073, "step": 97570 }, { "grad_norm": 0.11782202869653702, "learning_rate": 1.6015855169050355e-07, "loss": 0.0145, "step": 97580 }, { "grad_norm": 0.38817548751831055, "learning_rate": 1.5883890604543873e-07, "loss": 0.0108, "step": 97590 }, { "grad_norm": 0.3753117322921753, "learning_rate": 1.5752471095482057e-07, "loss": 0.0066, "step": 97600 }, { "grad_norm": 0.2856235206127167, "learning_rate": 1.5621596656235638e-07, "loss": 0.0122, "step": 97610 }, { "grad_norm": 0.28927838802337646, "learning_rate": 1.5491267301117608e-07, "loss": 0.0083, "step": 97620 }, { "grad_norm": 0.12534120678901672, "learning_rate": 1.536148304438101e-07, "loss": 0.0163, "step": 97630 }, { "grad_norm": 0.10946161299943924, "learning_rate": 1.5232243900217823e-07, "loss": 0.0105, "step": 97640 }, { "grad_norm": 0.06094136834144592, "learning_rate": 1.51035498827623e-07, "loss": 0.0103, "step": 97650 }, { "grad_norm": 0.4396219253540039, "learning_rate": 1.497540100608763e-07, "loss": 0.0135, "step": 97660 }, { "grad_norm": 0.038032349199056625, "learning_rate": 1.4847797284208153e-07, "loss": 0.016, "step": 97670 }, { "grad_norm": 0.09651615470647812, "learning_rate": 1.472073873107882e-07, "loss": 0.0139, "step": 97680 }, { "grad_norm": 0.1364695429801941, "learning_rate": 1.459422536059407e-07, "loss": 0.0133, "step": 97690 }, { "grad_norm": 0.7313283085823059, "learning_rate": 1.4468257186589508e-07, "loss": 0.0105, "step": 97700 }, { "grad_norm": 0.07506086677312851, "learning_rate": 1.4342834222840217e-07, "loss": 0.0117, "step": 97710 }, { "grad_norm": 0.2536759376525879, "learning_rate": 1.4217956483063566e-07, "loss": 0.0087, "step": 97720 }, { "grad_norm": 0.09941739588975906, "learning_rate": 1.409362398091474e-07, "loss": 0.0103, "step": 97730 }, { "grad_norm": 0.04819778725504875, "learning_rate": 1.3969836729990638e-07, "loss": 0.0083, "step": 97740 }, { "grad_norm": 0.1703205108642578, "learning_rate": 1.384659474382932e-07, "loss": 0.013, "step": 97750 }, { "grad_norm": 0.34355032444000244, "learning_rate": 1.372389803590779e-07, "loss": 0.0111, "step": 97760 }, { "grad_norm": 0.17539864778518677, "learning_rate": 1.360174661964364e-07, "loss": 0.0112, "step": 97770 }, { "grad_norm": 0.0437643900513649, "learning_rate": 1.3480140508396188e-07, "loss": 0.0104, "step": 97780 }, { "grad_norm": 0.12589310109615326, "learning_rate": 1.335907971546313e-07, "loss": 0.0083, "step": 97790 }, { "grad_norm": 0.12815260887145996, "learning_rate": 1.3238564254083875e-07, "loss": 0.0166, "step": 97800 }, { "grad_norm": 0.1331896334886551, "learning_rate": 1.3118594137437323e-07, "loss": 0.0129, "step": 97810 }, { "grad_norm": 0.19038663804531097, "learning_rate": 1.2999169378644094e-07, "loss": 0.007, "step": 97820 }, { "grad_norm": 0.29005831480026245, "learning_rate": 1.2880289990763738e-07, "loss": 0.0093, "step": 97830 }, { "grad_norm": 0.07675721496343613, "learning_rate": 1.2761955986796968e-07, "loss": 0.0074, "step": 97840 }, { "grad_norm": 0.1648579239845276, "learning_rate": 1.264416737968399e-07, "loss": 0.0129, "step": 97850 }, { "grad_norm": 0.07015048712491989, "learning_rate": 1.2526924182307275e-07, "loss": 0.0113, "step": 97860 }, { "grad_norm": 0.14475476741790771, "learning_rate": 1.2410226407487124e-07, "loss": 0.0114, "step": 97870 }, { "grad_norm": 0.26659196615219116, "learning_rate": 1.229407406798555e-07, "loss": 0.0097, "step": 97880 }, { "grad_norm": 0.013091398403048515, "learning_rate": 1.2178467176505726e-07, "loss": 0.007, "step": 97890 }, { "grad_norm": 0.26631712913513184, "learning_rate": 1.2063405745689205e-07, "loss": 0.0097, "step": 97900 }, { "grad_norm": 0.1407841145992279, "learning_rate": 1.1948889788119812e-07, "loss": 0.0093, "step": 97910 }, { "grad_norm": 0.1331661492586136, "learning_rate": 1.1834919316320303e-07, "loss": 0.0102, "step": 97920 }, { "grad_norm": 0.06555905938148499, "learning_rate": 1.1721494342754048e-07, "loss": 0.0081, "step": 97930 }, { "grad_norm": 0.10312109440565109, "learning_rate": 1.1608614879825563e-07, "loss": 0.0107, "step": 97940 }, { "grad_norm": 0.2490949034690857, "learning_rate": 1.1496280939879422e-07, "loss": 0.0124, "step": 97950 }, { "grad_norm": 0.19963957369327545, "learning_rate": 1.1384492535199687e-07, "loss": 0.0162, "step": 97960 }, { "grad_norm": 0.0907755121588707, "learning_rate": 1.1273249678011578e-07, "loss": 0.0123, "step": 97970 }, { "grad_norm": 0.16484642028808594, "learning_rate": 1.1162552380480362e-07, "loss": 0.0086, "step": 97980 }, { "grad_norm": 0.13351069390773773, "learning_rate": 1.1052400654711359e-07, "loss": 0.0118, "step": 97990 }, { "grad_norm": 0.04905006289482117, "learning_rate": 1.0942794512751597e-07, "loss": 0.011, "step": 98000 }, { "grad_norm": 0.14124324917793274, "learning_rate": 1.0833733966587045e-07, "loss": 0.015, "step": 98010 }, { "grad_norm": 0.14678896963596344, "learning_rate": 1.0725219028143718e-07, "loss": 0.0102, "step": 98020 }, { "grad_norm": 0.12471731752157211, "learning_rate": 1.0617249709289345e-07, "loss": 0.0094, "step": 98030 }, { "grad_norm": 0.07997379451990128, "learning_rate": 1.0509826021831703e-07, "loss": 0.0153, "step": 98040 }, { "grad_norm": 0.08212051540613174, "learning_rate": 1.0402947977517508e-07, "loss": 0.0116, "step": 98050 }, { "grad_norm": 0.18945151567459106, "learning_rate": 1.0296615588035185e-07, "loss": 0.0101, "step": 98060 }, { "grad_norm": 0.13296905159950256, "learning_rate": 1.0190828865012659e-07, "loss": 0.0106, "step": 98070 }, { "grad_norm": 0.10103214532136917, "learning_rate": 1.008558782001956e-07, "loss": 0.0114, "step": 98080 }, { "grad_norm": 0.40943023562431335, "learning_rate": 9.980892464563906e-08, "loss": 0.0158, "step": 98090 }, { "grad_norm": 0.21263021230697632, "learning_rate": 9.876742810095985e-08, "loss": 0.009, "step": 98100 }, { "grad_norm": 0.2830946147441864, "learning_rate": 9.77313886800446e-08, "loss": 0.0139, "step": 98110 }, { "grad_norm": 0.09697316586971283, "learning_rate": 9.670080649619717e-08, "loss": 0.0165, "step": 98120 }, { "grad_norm": 0.05471352860331535, "learning_rate": 9.567568166212737e-08, "loss": 0.009, "step": 98130 }, { "grad_norm": 0.1577693670988083, "learning_rate": 9.465601428992888e-08, "loss": 0.0091, "step": 98140 }, { "grad_norm": 0.3328542113304138, "learning_rate": 9.364180449111803e-08, "loss": 0.0146, "step": 98150 }, { "grad_norm": 0.04772527515888214, "learning_rate": 9.263305237661169e-08, "loss": 0.0161, "step": 98160 }, { "grad_norm": 0.19904109835624695, "learning_rate": 9.162975805671603e-08, "loss": 0.0087, "step": 98170 }, { "grad_norm": 0.037935130298137665, "learning_rate": 9.063192164115442e-08, "loss": 0.0088, "step": 98180 }, { "grad_norm": 0.10541808605194092, "learning_rate": 8.963954323904511e-08, "loss": 0.0083, "step": 98190 }, { "grad_norm": 0.08172605186700821, "learning_rate": 8.865262295891796e-08, "loss": 0.0135, "step": 98200 }, { "grad_norm": 0.18112370371818542, "learning_rate": 8.767116090870331e-08, "loss": 0.0064, "step": 98210 }, { "grad_norm": 0.1269218772649765, "learning_rate": 8.669515719572086e-08, "loss": 0.0097, "step": 98220 }, { "grad_norm": 0.3585854768753052, "learning_rate": 8.572461192671855e-08, "loss": 0.0118, "step": 98230 }, { "grad_norm": 0.258585125207901, "learning_rate": 8.475952520782815e-08, "loss": 0.0138, "step": 98240 }, { "grad_norm": 0.20307789742946625, "learning_rate": 8.379989714458747e-08, "loss": 0.0144, "step": 98250 }, { "grad_norm": 0.23395073413848877, "learning_rate": 8.284572784194034e-08, "loss": 0.0168, "step": 98260 }, { "grad_norm": 0.30715125799179077, "learning_rate": 8.189701740424216e-08, "loss": 0.0099, "step": 98270 }, { "grad_norm": 0.12410946190357208, "learning_rate": 8.095376593522663e-08, "loss": 0.0081, "step": 98280 }, { "grad_norm": 0.22768545150756836, "learning_rate": 8.001597353806123e-08, "loss": 0.0093, "step": 98290 }, { "grad_norm": 0.2820451855659485, "learning_rate": 7.90836403152917e-08, "loss": 0.0075, "step": 98300 }, { "grad_norm": 0.3922891318798065, "learning_rate": 7.815676636888092e-08, "loss": 0.0094, "step": 98310 }, { "grad_norm": 0.06343986093997955, "learning_rate": 7.723535180019226e-08, "loss": 0.005, "step": 98320 }, { "grad_norm": 0.14278095960617065, "learning_rate": 7.6319396709984e-08, "loss": 0.008, "step": 98330 }, { "grad_norm": 0.129511296749115, "learning_rate": 7.540890119843158e-08, "loss": 0.0106, "step": 98340 }, { "grad_norm": 0.02939336746931076, "learning_rate": 7.450386536509979e-08, "loss": 0.0119, "step": 98350 }, { "grad_norm": 0.2548007071018219, "learning_rate": 7.360428930895947e-08, "loss": 0.0157, "step": 98360 }, { "grad_norm": 0.11397193372249603, "learning_rate": 7.271017312839302e-08, "loss": 0.0173, "step": 98370 }, { "grad_norm": 0.044094718992710114, "learning_rate": 7.182151692118333e-08, "loss": 0.0178, "step": 98380 }, { "grad_norm": 0.18133437633514404, "learning_rate": 7.093832078449713e-08, "loss": 0.0124, "step": 98390 }, { "grad_norm": 0.1580927073955536, "learning_rate": 7.006058481493493e-08, "loss": 0.0087, "step": 98400 }, { "grad_norm": 0.1607571542263031, "learning_rate": 6.918830910847552e-08, "loss": 0.0122, "step": 98410 }, { "grad_norm": 0.46527785062789917, "learning_rate": 6.832149376052033e-08, "loss": 0.0088, "step": 98420 }, { "grad_norm": 0.20655693113803864, "learning_rate": 6.746013886584912e-08, "loss": 0.0164, "step": 98430 }, { "grad_norm": 0.2271413654088974, "learning_rate": 6.660424451866431e-08, "loss": 0.0122, "step": 98440 }, { "grad_norm": 0.24005912244319916, "learning_rate": 6.575381081256327e-08, "loss": 0.0101, "step": 98450 }, { "grad_norm": 0.3067402243614197, "learning_rate": 6.49088378405549e-08, "loss": 0.0143, "step": 98460 }, { "grad_norm": 0.1611158698797226, "learning_rate": 6.406932569503755e-08, "loss": 0.0172, "step": 98470 }, { "grad_norm": 0.08098594844341278, "learning_rate": 6.323527446782662e-08, "loss": 0.0122, "step": 98480 }, { "grad_norm": 0.08147673308849335, "learning_rate": 6.240668425012142e-08, "loss": 0.0098, "step": 98490 }, { "grad_norm": 0.3463030755519867, "learning_rate": 6.158355513254388e-08, "loss": 0.0151, "step": 98500 }, { "grad_norm": 0.4162518382072449, "learning_rate": 6.076588720510534e-08, "loss": 0.0094, "step": 98510 }, { "grad_norm": 0.16584697365760803, "learning_rate": 5.995368055722316e-08, "loss": 0.0137, "step": 98520 }, { "grad_norm": 0.05641373619437218, "learning_rate": 5.914693527773185e-08, "loss": 0.012, "step": 98530 }, { "grad_norm": 0.2743690609931946, "learning_rate": 5.834565145483861e-08, "loss": 0.0089, "step": 98540 }, { "grad_norm": 0.28121453523635864, "learning_rate": 5.754982917618446e-08, "loss": 0.0167, "step": 98550 }, { "grad_norm": 0.06935794651508331, "learning_rate": 5.675946852878866e-08, "loss": 0.0086, "step": 98560 }, { "grad_norm": 0.02581387385725975, "learning_rate": 5.59745695990932e-08, "loss": 0.0123, "step": 98570 }, { "grad_norm": 0.0794103741645813, "learning_rate": 5.51951324729294e-08, "loss": 0.0104, "step": 98580 }, { "grad_norm": 0.1604519486427307, "learning_rate": 5.4421157235529093e-08, "loss": 0.0067, "step": 98590 }, { "grad_norm": 0.17258961498737335, "learning_rate": 5.365264397154679e-08, "loss": 0.0122, "step": 98600 }, { "grad_norm": 0.0581522211432457, "learning_rate": 5.288959276501526e-08, "loss": 0.0132, "step": 98610 }, { "grad_norm": 0.18455199897289276, "learning_rate": 5.2132003699378876e-08, "loss": 0.0104, "step": 98620 }, { "grad_norm": 0.12216533720493317, "learning_rate": 5.137987685749912e-08, "loss": 0.0088, "step": 98630 }, { "grad_norm": 0.1160929724574089, "learning_rate": 5.0633212321610227e-08, "loss": 0.0097, "step": 98640 }, { "grad_norm": 0.044484857469797134, "learning_rate": 4.989201017338019e-08, "loss": 0.0097, "step": 98650 }, { "grad_norm": 0.1001085564494133, "learning_rate": 4.9156270493860846e-08, "loss": 0.0109, "step": 98660 }, { "grad_norm": 0.175063356757164, "learning_rate": 4.842599336351561e-08, "loss": 0.0123, "step": 98670 }, { "grad_norm": 0.20719614624977112, "learning_rate": 4.7701178862197274e-08, "loss": 0.0092, "step": 98680 }, { "grad_norm": 0.098584845662117, "learning_rate": 4.698182706918131e-08, "loss": 0.0103, "step": 98690 }, { "grad_norm": 0.28288504481315613, "learning_rate": 4.6267938063121466e-08, "loss": 0.0128, "step": 98700 }, { "grad_norm": 0.03299727663397789, "learning_rate": 4.555951192209973e-08, "loss": 0.0103, "step": 98710 }, { "grad_norm": 0.3263688087463379, "learning_rate": 4.485654872358747e-08, "loss": 0.0084, "step": 98720 }, { "grad_norm": 0.1154012680053711, "learning_rate": 4.415904854445097e-08, "loss": 0.0117, "step": 98730 }, { "grad_norm": 0.3606019616127014, "learning_rate": 4.34670114609792e-08, "loss": 0.0149, "step": 98740 }, { "grad_norm": 0.26227885484695435, "learning_rate": 4.278043754884498e-08, "loss": 0.0108, "step": 98750 }, { "grad_norm": 0.44846880435943604, "learning_rate": 4.209932688312712e-08, "loss": 0.0141, "step": 98760 }, { "grad_norm": 0.45740029215812683, "learning_rate": 4.14236795383216e-08, "loss": 0.0156, "step": 98770 }, { "grad_norm": 0.17394967377185822, "learning_rate": 4.075349558830821e-08, "loss": 0.0139, "step": 98780 }, { "grad_norm": 0.3047555387020111, "learning_rate": 4.008877510638387e-08, "loss": 0.0124, "step": 98790 }, { "grad_norm": 0.28516823053359985, "learning_rate": 3.942951816523488e-08, "loss": 0.0115, "step": 98800 }, { "grad_norm": 0.09110671281814575, "learning_rate": 3.8775724836959125e-08, "loss": 0.0108, "step": 98810 }, { "grad_norm": 0.02978845313191414, "learning_rate": 3.812739519305497e-08, "loss": 0.0066, "step": 98820 }, { "grad_norm": 0.11559776961803436, "learning_rate": 3.748452930442126e-08, "loss": 0.0074, "step": 98830 }, { "grad_norm": 0.0742420107126236, "learning_rate": 3.6847127241362855e-08, "loss": 0.0065, "step": 98840 }, { "grad_norm": 0.08605418354272842, "learning_rate": 3.621518907358512e-08, "loss": 0.0135, "step": 98850 }, { "grad_norm": 0.050393685698509216, "learning_rate": 3.558871487019388e-08, "loss": 0.0113, "step": 98860 }, { "grad_norm": 0.06940548121929169, "learning_rate": 3.4967704699701007e-08, "loss": 0.0139, "step": 98870 }, { "grad_norm": 0.08390871435403824, "learning_rate": 3.435215863001884e-08, "loss": 0.0084, "step": 98880 }, { "grad_norm": 0.04726475849747658, "learning_rate": 3.374207672846019e-08, "loss": 0.015, "step": 98890 }, { "grad_norm": 0.48558005690574646, "learning_rate": 3.313745906174392e-08, "loss": 0.0138, "step": 98900 }, { "grad_norm": 0.040919750928878784, "learning_rate": 3.253830569599492e-08, "loss": 0.0093, "step": 98910 }, { "grad_norm": 0.09018481522798538, "learning_rate": 3.1944616696727436e-08, "loss": 0.0079, "step": 98920 }, { "grad_norm": 0.09400053322315216, "learning_rate": 3.135639212887287e-08, "loss": 0.0058, "step": 98930 }, { "grad_norm": 0.06361251324415207, "learning_rate": 3.077363205675754e-08, "loss": 0.0078, "step": 98940 }, { "grad_norm": 0.18683703243732452, "learning_rate": 3.0196336544113804e-08, "loss": 0.0129, "step": 98950 }, { "grad_norm": 0.11915512382984161, "learning_rate": 2.9624505654063383e-08, "loss": 0.0126, "step": 98960 }, { "grad_norm": 0.14803628623485565, "learning_rate": 2.90581394491507e-08, "loss": 0.0098, "step": 98970 }, { "grad_norm": 0.07482381910085678, "learning_rate": 2.8497237991309545e-08, "loss": 0.0089, "step": 98980 }, { "grad_norm": 0.3702681362628937, "learning_rate": 2.7941801341879735e-08, "loss": 0.0123, "step": 98990 }, { "grad_norm": 0.08928745985031128, "learning_rate": 2.7391829561601578e-08, "loss": 0.0077, "step": 99000 }, { "grad_norm": 0.08280210196971893, "learning_rate": 2.6847322710621403e-08, "loss": 0.008, "step": 99010 }, { "grad_norm": 0.06509878486394882, "learning_rate": 2.6308280848486022e-08, "loss": 0.007, "step": 99020 }, { "grad_norm": 0.08241214603185654, "learning_rate": 2.5774704034137177e-08, "loss": 0.0137, "step": 99030 }, { "grad_norm": 0.1702146679162979, "learning_rate": 2.524659232593929e-08, "loss": 0.0102, "step": 99040 }, { "grad_norm": 0.2372375726699829, "learning_rate": 2.472394578163506e-08, "loss": 0.0125, "step": 99050 }, { "grad_norm": 0.03384452313184738, "learning_rate": 2.4206764458378772e-08, "loss": 0.0098, "step": 99060 }, { "grad_norm": 0.08745351433753967, "learning_rate": 2.3695048412736285e-08, "loss": 0.0061, "step": 99070 }, { "grad_norm": 0.25894758105278015, "learning_rate": 2.3188797700668395e-08, "loss": 0.0091, "step": 99080 }, { "grad_norm": 0.5202571749687195, "learning_rate": 2.268801237753082e-08, "loss": 0.0144, "step": 99090 }, { "grad_norm": 0.3355400562286377, "learning_rate": 2.2192692498090862e-08, "loss": 0.0098, "step": 99100 }, { "grad_norm": 0.1310136318206787, "learning_rate": 2.170283811652185e-08, "loss": 0.0086, "step": 99110 }, { "grad_norm": 0.3630449175834656, "learning_rate": 2.12184492863865e-08, "loss": 0.0138, "step": 99120 }, { "grad_norm": 0.039683934301137924, "learning_rate": 2.073952606066465e-08, "loss": 0.0083, "step": 99130 }, { "grad_norm": 0.5092173218727112, "learning_rate": 2.026606849171997e-08, "loss": 0.0094, "step": 99140 }, { "grad_norm": 0.1421227753162384, "learning_rate": 1.9798076631333262e-08, "loss": 0.0181, "step": 99150 }, { "grad_norm": 0.11193552613258362, "learning_rate": 1.933555053069136e-08, "loss": 0.0115, "step": 99160 }, { "grad_norm": 0.2993013858795166, "learning_rate": 1.887849024036492e-08, "loss": 0.012, "step": 99170 }, { "grad_norm": 0.031051361933350563, "learning_rate": 1.8426895810341736e-08, "loss": 0.0087, "step": 99180 }, { "grad_norm": 0.3059634566307068, "learning_rate": 1.798076729000453e-08, "loss": 0.0094, "step": 99190 }, { "grad_norm": 0.2571016252040863, "learning_rate": 1.75401047281476e-08, "loss": 0.0119, "step": 99200 }, { "grad_norm": 0.05541766434907913, "learning_rate": 1.7104908172954626e-08, "loss": 0.0094, "step": 99210 }, { "grad_norm": 0.11934835463762283, "learning_rate": 1.6675177672020868e-08, "loss": 0.0156, "step": 99220 }, { "grad_norm": 0.0314561128616333, "learning_rate": 1.6250913272342073e-08, "loss": 0.0053, "step": 99230 }, { "grad_norm": 0.26054275035858154, "learning_rate": 1.5832115020314453e-08, "loss": 0.0149, "step": 99240 }, { "grad_norm": 0.05562305077910423, "learning_rate": 1.5418782961734712e-08, "loss": 0.0078, "step": 99250 }, { "grad_norm": 0.07863426208496094, "learning_rate": 1.501091714181113e-08, "loss": 0.0061, "step": 99260 }, { "grad_norm": 0.11702148616313934, "learning_rate": 1.460851760513582e-08, "loss": 0.0084, "step": 99270 }, { "grad_norm": 0.1265617460012436, "learning_rate": 1.4211584395723566e-08, "loss": 0.007, "step": 99280 }, { "grad_norm": 0.23976677656173706, "learning_rate": 1.382011755697854e-08, "loss": 0.007, "step": 99290 }, { "grad_norm": 0.17295686900615692, "learning_rate": 1.3434117131716494e-08, "loss": 0.0064, "step": 99300 }, { "grad_norm": 0.03503004088997841, "learning_rate": 1.3053583162142557e-08, "loss": 0.0086, "step": 99310 }, { "grad_norm": 0.21571411192417145, "learning_rate": 1.2678515689873438e-08, "loss": 0.0099, "step": 99320 }, { "grad_norm": 0.15050512552261353, "learning_rate": 1.2308914755931877e-08, "loss": 0.0128, "step": 99330 }, { "grad_norm": 0.10995852947235107, "learning_rate": 1.1944780400729993e-08, "loss": 0.0135, "step": 99340 }, { "grad_norm": 0.13528850674629211, "learning_rate": 1.1586112664085935e-08, "loss": 0.0131, "step": 99350 }, { "grad_norm": 0.09705416858196259, "learning_rate": 1.1232911585234984e-08, "loss": 0.0106, "step": 99360 }, { "grad_norm": 0.4191369116306305, "learning_rate": 1.0885177202790697e-08, "loss": 0.0163, "step": 99370 }, { "grad_norm": 0.21940429508686066, "learning_rate": 1.0542909554789315e-08, "loss": 0.009, "step": 99380 }, { "grad_norm": 0.18792329728603363, "learning_rate": 1.0206108678656456e-08, "loss": 0.0104, "step": 99390 }, { "grad_norm": 0.10130082070827484, "learning_rate": 9.874774611223769e-09, "loss": 0.0107, "step": 99400 }, { "grad_norm": 0.05188845843076706, "learning_rate": 9.548907388728933e-09, "loss": 0.0072, "step": 99410 }, { "grad_norm": 0.35063424706459045, "learning_rate": 9.228507046804557e-09, "loss": 0.0127, "step": 99420 }, { "grad_norm": 0.05222804844379425, "learning_rate": 8.91357362048928e-09, "loss": 0.0109, "step": 99430 }, { "grad_norm": 0.16175979375839233, "learning_rate": 8.604107144227769e-09, "loss": 0.0118, "step": 99440 }, { "grad_norm": 0.39839908480644226, "learning_rate": 8.300107651859623e-09, "loss": 0.009, "step": 99450 }, { "grad_norm": 0.03716075420379639, "learning_rate": 8.001575176630472e-09, "loss": 0.0092, "step": 99460 }, { "grad_norm": 0.10818666219711304, "learning_rate": 7.708509751186422e-09, "loss": 0.0118, "step": 99470 }, { "grad_norm": 0.176832914352417, "learning_rate": 7.420911407579611e-09, "loss": 0.0141, "step": 99480 }, { "grad_norm": 0.3794540762901306, "learning_rate": 7.13878017725711e-09, "loss": 0.014, "step": 99490 }, { "grad_norm": 0.16482484340667725, "learning_rate": 6.8621160910720125e-09, "loss": 0.0088, "step": 99500 }, { "grad_norm": 0.10975614935159683, "learning_rate": 6.59091917928345e-09, "loss": 0.0078, "step": 99510 }, { "grad_norm": 0.2477596402168274, "learning_rate": 6.32518947155103e-09, "loss": 0.0141, "step": 99520 }, { "grad_norm": 0.07513676583766937, "learning_rate": 6.064926996929288e-09, "loss": 0.0119, "step": 99530 }, { "grad_norm": 0.04990684241056442, "learning_rate": 5.810131783884343e-09, "loss": 0.0092, "step": 99540 }, { "grad_norm": 0.1508452296257019, "learning_rate": 5.5608038602772415e-09, "loss": 0.0143, "step": 99550 }, { "grad_norm": 0.07930289953947067, "learning_rate": 5.316943253375062e-09, "loss": 0.0109, "step": 99560 }, { "grad_norm": 0.04715585336089134, "learning_rate": 5.07854998984536e-09, "loss": 0.0088, "step": 99570 }, { "grad_norm": 0.25370457768440247, "learning_rate": 4.845624095756174e-09, "loss": 0.0122, "step": 99580 }, { "grad_norm": 0.3418320417404175, "learning_rate": 4.618165596587121e-09, "loss": 0.0133, "step": 99590 }, { "grad_norm": 0.12892672419548035, "learning_rate": 4.396174517207197e-09, "loss": 0.0134, "step": 99600 }, { "grad_norm": 0.14365698397159576, "learning_rate": 4.179650881896979e-09, "loss": 0.0125, "step": 99610 }, { "grad_norm": 0.059959776699543, "learning_rate": 3.968594714331975e-09, "loss": 0.0076, "step": 99620 }, { "grad_norm": 0.1423734873533249, "learning_rate": 3.763006037593719e-09, "loss": 0.0121, "step": 99630 }, { "grad_norm": 0.2244083434343338, "learning_rate": 3.5628848741586786e-09, "loss": 0.0051, "step": 99640 }, { "grad_norm": 0.15286479890346527, "learning_rate": 3.368231245926001e-09, "loss": 0.0138, "step": 99650 }, { "grad_norm": 0.1547263115644455, "learning_rate": 3.179045174167561e-09, "loss": 0.0092, "step": 99660 }, { "grad_norm": 0.07790018618106842, "learning_rate": 2.9953266795834657e-09, "loss": 0.0109, "step": 99670 }, { "grad_norm": 0.06365662068128586, "learning_rate": 2.817075782263201e-09, "loss": 0.0119, "step": 99680 }, { "grad_norm": 0.30339738726615906, "learning_rate": 2.6442925016967325e-09, "loss": 0.0124, "step": 99690 }, { "grad_norm": 0.034184761345386505, "learning_rate": 2.4769768567800555e-09, "loss": 0.0076, "step": 99700 }, { "grad_norm": 0.18588736653327942, "learning_rate": 2.315128865809646e-09, "loss": 0.0106, "step": 99710 }, { "grad_norm": 0.4204089045524597, "learning_rate": 2.1587485464880097e-09, "loss": 0.0167, "step": 99720 }, { "grad_norm": 0.12317453324794769, "learning_rate": 2.0078359159125813e-09, "loss": 0.0104, "step": 99730 }, { "grad_norm": 0.18558384478092194, "learning_rate": 1.8623909905923776e-09, "loss": 0.0156, "step": 99740 }, { "grad_norm": 0.02937978506088257, "learning_rate": 1.7224137864257916e-09, "loss": 0.0124, "step": 99750 }, { "grad_norm": 0.12174257636070251, "learning_rate": 1.5879043187283505e-09, "loss": 0.0135, "step": 99760 }, { "grad_norm": 0.12686564028263092, "learning_rate": 1.458862602204958e-09, "loss": 0.0122, "step": 99770 }, { "grad_norm": 0.11697624623775482, "learning_rate": 1.3352886509720996e-09, "loss": 0.014, "step": 99780 }, { "grad_norm": 0.07793644815683365, "learning_rate": 1.2171824785356389e-09, "loss": 0.009, "step": 99790 }, { "grad_norm": 0.1921953707933426, "learning_rate": 1.1045440978185718e-09, "loss": 0.0106, "step": 99800 }, { "grad_norm": 0.2720211148262024, "learning_rate": 9.973735211332714e-10, "loss": 0.0138, "step": 99810 }, { "grad_norm": 0.2894802391529083, "learning_rate": 8.956707602036929e-10, "loss": 0.0078, "step": 99820 }, { "grad_norm": 0.04237085580825806, "learning_rate": 7.994358261542712e-10, "loss": 0.0057, "step": 99830 }, { "grad_norm": 0.18995033204555511, "learning_rate": 7.086687295043692e-10, "loss": 0.0097, "step": 99840 }, { "grad_norm": 0.20793896913528442, "learning_rate": 6.23369480179381e-10, "loss": 0.0104, "step": 99850 }, { "grad_norm": 0.08254154026508331, "learning_rate": 5.43538087510731e-10, "loss": 0.0098, "step": 99860 }, { "grad_norm": 0.5427705645561218, "learning_rate": 4.691745602303233e-10, "loss": 0.0304, "step": 99870 }, { "grad_norm": 0.045003436505794525, "learning_rate": 4.0027890646499034e-10, "loss": 0.0078, "step": 99880 }, { "grad_norm": 0.07736022770404816, "learning_rate": 3.368511337531466e-10, "loss": 0.0083, "step": 99890 }, { "grad_norm": 0.26510486006736755, "learning_rate": 2.788912490281348e-10, "loss": 0.0071, "step": 99900 }, { "grad_norm": 0.18111801147460938, "learning_rate": 2.2639925863487954e-10, "loss": 0.0082, "step": 99910 }, { "grad_norm": 0.3267061710357666, "learning_rate": 1.7937516830213163e-10, "loss": 0.0098, "step": 99920 }, { "grad_norm": 0.3035076856613159, "learning_rate": 1.3781898318687702e-10, "loss": 0.0105, "step": 99930 }, { "grad_norm": 0.10614597052335739, "learning_rate": 1.0173070781882566e-10, "loss": 0.0072, "step": 99940 }, { "grad_norm": 0.10206140577793121, "learning_rate": 7.111034615592261e-11, "loss": 0.0107, "step": 99950 }, { "grad_norm": 0.538131058216095, "learning_rate": 4.595790153993918e-11, "loss": 0.0106, "step": 99960 }, { "grad_norm": 0.11234001815319061, "learning_rate": 2.6273376729779585e-11, "loss": 0.0089, "step": 99970 }, { "grad_norm": 0.08032608032226562, "learning_rate": 1.2056773868174276e-11, "loss": 0.0118, "step": 99980 }, { "grad_norm": 0.19018392264842987, "learning_rate": 3.3080945149865926e-12, "loss": 0.0093, "step": 99990 }, { "grad_norm": 0.19717437028884888, "learning_rate": 2.7339625008337977e-14, "loss": 0.0197, "step": 100000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 2000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }